Patients' Demographic Features
In this study, we prospectively included a total of 737 IgG4-RD patients divided into five groups according to age at diagnosis (Figure 1A). Male ratio increased with advancing age (Figure 1B, P for trend <0.001). In addition, the demographic features are shown in Table 1. The disease duration (the time from onset to diagnosis) and frequencies of allergy history were significantly different across age groups. The number of organs involved was comparable in different age groups. At disease onset, the occurrence of lacrimal gland swelling (P < 0.001), nausea and vomiting (P = 0.028), jaundice (P < 0.001) and nasal congestion (P < 0.001) were significantly different across age groups.
Age, sex and organ involvement distributions of IgG4-RD patients
(A) Age distributions of patients. (B) Sex ratio of patients across age groups. (C) Heat map: the percentages of organ involvement across age groups. (D–F) Comparison of percentages of total superficial organs (D), lacrimal gland (E) and sinus (F) involvement across age groups. (G–I) Comparison of percentages of total internal organs (G), pancreas (H) and biliary tract (I) involvement across age groups. GT: gastrointestinal tract; SG: submandibular gland.
In terms of organs involved in IgG4-RD, a heat map was used to elucidate and visualize the difference across age groups (Figure 1C), showing different frequencies of organ involvement in the age groups. To further demonstrate the characteristics of organ involvement based on age disparities, we subsequently compared the proportions of superficial or internal organ involvement and analysed the change patterns across age groups. The proportions of total superficial organs (P for trend <0.001), lacrimal gland (P for trend <0.001) and sinus (P for trend <0.001) involvement were significantly different between age groups, all presenting downtrends with advancing age (Figure 1D–F). However, the proportions of other superficial organ involvement were comparable in different age groups (Supplementary Figure S1, available at Rheumatology online). In addition, the involvement rates of total internal organs (P for trend <0.001), pancreas (P for trend <0.001), biliary tract (P for trend <0.001), lung (P for trend =0.022), prostate (P for trend =0.001) and retroperitoneum (P for trend <0.001) were significantly different across age groups, presenting uptrends with advancing age (Figure 1G–I and Supplementary Figure S1, available at Rheumatology online). In other internal organs, the percentage involvement was not significantly different across age groups (Supplementary Figure S1, available at Rheumatology online). Considering that female patients were more likely to present with superficial organ involvement, while male patients more frequently had internal organ involvement, we wanted to know whether this age-related organ involvement disparity was affected by the distribution of gender. Hence, we compared the organ involvements across age groups in male and female (Supplementary Figure S2, available at Rheumatology online). The proportions of superficial organ (lacrimal gland and sinus) involvement decreased with age in both male and female patients. On the contrary, the proportion of internal organ involvement increased with age, and was more prominent in male patient, mainly presenting in pancreas, biliary tract, lung and retroperitoneum, while the phenomenon only occurred in biliary tract in female patients.
We further compared the laboratory tests at baseline among the five age groups. Haemoglobin (P =0.028), platelet (P < 0.001), ESR (P < 0.001), CRP (P = 0.012), serum IgG (P = 0.001), IgM (P < 0.001), IgG1 (P < 0.001), IgG3 (P < 0.001), IgG4 (P = 0.006) concentrations and proportions of B regulatory cells (CD19+CD24hiCD38hi) (P = 0.038) were significantly different across age groups (Table 2).
We also performed correlation analysis between laboratory tests and age (Figure 2). We found that ESR (r = 0.263, P < 0.001), CRP (r = 0.175, P < 0.001), serum IgG (r = 0.13, P < 0.001), IgG1 (r = 0.168, P < 0.001), IgG3 (r = 0.165, P < 0.001) and IgG4 concentration (r = 0.106, P = 0.004) were positively correlated with age. What is more, serum IgM concentration (r = −0.184, P < 0.001) and proportions of naïve B cell (CD19+CD24intCD38int) (r = −0.202, P = 0.043) were negatively correlated with age. However, none of peripheral eosinophils, serum IgA, T-IgE, IgG2 concentrations or other B cell subsets showed statistically significant correlation with age (data not shown).
Correlation analysis between age and baseline laboratory findings
(A and B) Correlations between ESR/CRP and age. (C–G) Correlations between serum levels of IgG/M/G1/G3/G4 and age. (H) Correlation between proportions of CD19+CD24intCD38int cells in CD19+ cells and age.
Line charts were used to describe the patterns of change of serum T-IgE and IgG subclasses with age (Figure 3A). Besides, we also recorded the changes of serum IgG4 level at 3, 6 and 12 months after treatment (Figure 3B). The serum IgG1, IgG3, IgG4 and T-IgE levels peaked at 60–69 years of age and serum IgG4 level in the age group 60–69 years remained at a peak after treatment for 3, 6 and 12 months.
Levels of IgG subclasses and T-IgE, IgG4-RD RI, treatment approaches and treatment outcomes across age groups
(A) Median values of serum IgG subclasses and T-IgE across age groups. (B) Median values of serum IgG4 at baseline or after treatment for 3, 6 and 12 months. (C) Boxplots of IgG4-RD RI across age groups. (D) Percentage of treatment approaches across age groups. (E) Kaplan–Meier survival curve of patients without relapse between young patients (age ≤56) and elderly patients (age >56). IgG4-RD: IgG4-related disease; RI: responder index; T-IgE: total IgE.
Furthermore, multivariate linear regression analysis was performed to identify the potential factors associated with baseline serum IgG4 level (Supplementary Table S1, available at Rheumatology online). As shown in Supplementary Table S1, age was independently associated with serum IgG4 level (standardized β-coefficient 0.146, P < 0.001). Other identified variables were sex (standardized β-coefficient 0.161, P < 0.001), lacrimal gland involvement (standardized β-coefficient 0.226, P < 0.001), artery involvement (standardized β-coefficient −0.15, P < 0.001) and number of involved organs (standardized β-coefficient 0.332, P < 0.001).
Disease Activity, Treatment Agents and Outcomes
We further compared the IgG4-RD RI of patients among the five groups (Figure 3C) and it showed a significant difference in age groups (P = 0.025). The treatment approaches were not significantly different (Figure 3D, P = 0.39). A total of 310 patients who were treated with GC alone or GC combined with IM were included in the study of treatment outcomes and followed up for >36 months. The receiver operating characteristic curve revealed that age ≤56.5 was strongly linked to disease relapse. As a result, patients were divided into young patients (age ≤56, n = 173) and elderly patients (age >56, n = 137). Using PSM, 119 patients from the young patients and 119 patients from the elderly patients were matched. Baseline clinical characteristics of the two groups were balanced after PSM (Supplementary Table S2, available at Rheumatology online). A total of 53 patients (22.3%) relapsed during the follow-up of 36 months and Kaplan–Meier survival analysis revealed that young patients (age ≤56) were more likely to relapse over any given month of follow-up compared with elderly patients (Figure 3E, P = 0.029).
Multivariate Cox analysis of 310 patients treated with GC alone or GC combined with IM identified that older age (hazard ratio 0.54, P = 0.013) was the independent protective factor with relapse, while higher percentage of eosinophils (hazard ratio 2.44, P < 0.001) was the independent risk factor with relapse (Table 3).
Clinical Characteristics and Treatment Outcomes in Paediatric Patients
We also studied the clinical characteristics and treatment outcomes of paediatric patients (Supplementary Table S3, available at Rheumatology online). Ten patients (seven boys and three girls) aged from 9 to 17 years were included. Disease duration ranged from 4 to 60 months. Most patients presented with Mikulicz's disease (70%). Other manifestations were lymphadenopathy (40%), autoimmune pancreatitis (30%), IgG4-related sclerosing cholangitis (20%), lung disease (10%), sinusitis (10%), hepatopathy (10%) and thyroiditis (10%). Multiple organ involvement (two or more organ manifestations) occurred in 60% of paediatric patients. The percentage of the paediatric patients with serum IgG4 elevation was significantly lower than that in adults (70% vs 92.6%, P = 0.036). The IgG4-RD RI ranged from 3 to 21. The majority of patients (70%) were treated with GC alone or combined with IM. In the seven patients treated with GC monotherapy or GC combined with IM, three patients (42.9%) relapsed in the follow-up.
Rheumatology. 2021;60(6):2635-2646. © 2021 Oxford University Press