Demographic characteristics of IgG4-RD with hypocomplementemia
In this study, we prospectively enrolled 312 newly diagnosed IgG4-RD patients, 65(20.83%) patients had hypocomplementemia (hypocomplementemia group), 244(78.2%) patients had normal complement (normal complement group) and 3(1.0%) patients had elevated complement. Of the hypocomplementemia group, 45(69.2%) cases had both complement C3 and C4 reduction, 14(21.5%) cases with only C3 reduction, 6(9.2%) cases with only C4 reduction. As the number of cases with elevated complement was very small, we mainly compared and discussed hypocomplementemia group and normal complement group. Demographic features of such two groups were shown in Table 1. The age at diagnose in hypocomplementemia patients was 55.85±10.89 years, higher than normal complement group. The median duration of disease prior to initial evaluation was 12(4, 36) months. There was no significant difference of incidence of allergic history between two groups. Compared with normal complement group, patients with hypocomplementemia showed significantly male predominance (72.3% vs 58.6%, P=0.044), more number of involved organs (4.88±1.79 vs 2.89±1.36, P<0.001) and higher IgG4-RD RI (15.74±5.78 vs 9.64±4.33, P<0.001) at baseline.
Comparison of involved organs in hypocomplementemia group and normal complement group
Our data demonstrate the discrepancies in the clinical spectrums between two groups. Compared with normal complement group,patients with hypocomplementemia had significantly higher incidence of lymph node (66.2% vs 36.1%, P<0.001), lacrimal gland (66.2% vs 45.5%, P=0.003), submandibular gland (63.1% vs 41.4%, P=0.001), pancreas (50.8% vs 27.1%, P<0.001), lung (50.8% vs 18.0%, P<0.001), paranasal sinus (41.5% vs 27.9% P=0.029), parotid gland (33.8% vs 11.5%,, P<0.001), bile duct (30.8% vs 14.3%, P=0.002) and prostate gland (15.4% vs 4.1%, P =0.021) (Table 1). There was no significant difference in kidney involvement between the two groups.
Comparison of laboratory parameters in IgG4-RD hypocomplementemia group and normal complement group
The average level of serum C3 in hypocomplementemia was 0.54±0.17g/L (normal 0.73-1.46 g/L) and C4 was 0.061±0.047 g/L(normal 0.10-0.40g/L)). We further compared the laboratory tests between two groups (Table1.), and found that patients with hypocomplementemia had significantly higher baseline levels of peripheral eosinophils count(median 0.42´109/L vs 0.17´109/L, P=0.006), ESR (46.34±32.40mm/h vs median 16mm/h, P<0.001), IgG (30.92±15.31 g/L vs 18.05±8.79 g/L, P<0.001), total-IgE (median 471.0 KU/L vs 222.0 KU/L, P<0.001), IgG1 (1295.11±539.48mg/dL vs 907.73±439.79mg/dL, P<0.001), IgG3 (100.56±80.81mg/dL vs 50.54±41.23mg/dL, P<0.001), IgG4 (2614.13±1915.39mg/dL vs median 547.50mg/dL, P<0.001 ), ratio of IgG4/IgG(0.80±0.44 vs 0.44±0.34, P<0.001 ), RF positive rate (42.6% vs 26.5%, P=0.034 ), whereas significantly lower count of platelet(215.80±59.00´109/L vs 240.30±75.80´109/L, P=0.017), IgA (1.59±0.78g/L vs 2.36±1.32g/L, P<0.001).
Correlations between serum C3, C4 and clinical characteristics at baseline
We performed Pearson correlation coefficient analysis to investigate the association between serum complement level and age of onset, duration of disease, number of involved organs, IgG4-RD RI and laboratory parameters including C3/C4, count of eosinophil, ESR, hsCRP, IgG, IgA, IgM, IgE, IgG1, IgG2, IgG3, IgG4, in total patients at baseline. As shown in figure 1, level of serum C3 was negatively correlated with the age (r=-0.162, P=0.005 ), number of involved organs (r=-0.441, P<0.001), IgG4-RD RI (r=-0.201, P=0.005),IgG (r=-0.362, P<0.001), IgG1(r=-0.216, P<0.001), IgG3 (r=-0.338, P<0.001) and IgG4 (r=-0.425, P<0.001), whereas positively correlated with serum C4(r=0.726, P<0.001), hsCRP (r=0.203, P=0.002) and IgA (r=0.341, P<0.001). Similarly, serum C4 level was negatively correlated with number of involved organs (r=-0.309, P<0.001), IgG4-RD RI (r=-0.207, P<0.001),laboratory parameters such as IgG (r=-0.436, P<0.001), IgG1(r=-0.315, P<0.001), IgG3 (r=-0.301, P<0.001), IgG4 (r=-0.422, P<0.001), positively correlated with serum C3(r=0.726, P<0.001), hsCRP (r=0.192, P=0.003), IgA (r=0.224, P<0.001).
Treatment efficacy in IgG4-RD with hypocomplementemia
All patients with hypocomplementemia were treated with glucocorticoids (GCs), GCs combined with immunosuppressant agents (GCs plus IM) or GCs combined with rituximab (RTX). The standard induction dosage of oral prednisone was 0.6~1.0 mg kg-1day-1 in the first month and tapered per 1 or 2 weeks to the maintenance dosage.18 (27.7%) patients received GC monotherapy. One (1.5%) patient received GCs plus RTX. Others were treated with GCs plus IM, including cyclophosphamide (CYC) (n=22, 33.8%), mycophenolate mofetil (MMF) (n=12, 18.5%), methotrexate (MTX) (n=5, 7.7%), iguratimod 5(n=5, 7.7%) and leflunomide (n=1, 1.5%).
The average follow-up time of IgG4-RD patients with hypocomplementemia was 34.01±18.34 months. Level of serum C3 and C4 increased to the normal range the first month after treatment (Fig2A, B). Laboratory parameters such as ESR, hsCRP, IgG (Fig2C), IgG1 (Fig2D), IgG4 (Fig2E), IgE (Fig2F) decreased significantly after treatment. Disease response occurred in 64(98.5%) patients at month 3 and were observed quickly. One patient had no improvement at the 3rd month until increased the dosage of GCs and combined with IMs. Fifteen (23.1%) patients relapsed during follow-up with mean recurrence time 14.2±13.8 months,while only 25% of them had hypocomplementemia while relapsed.
Comparison of treatment and outcome in IgG4-RD hypocomplementemia group and normal complement group
We compared the first-line treatment between the two groups. The initial doses of GCs in hypocomplementemia group were higher than the normal complement group (37.77±14.28 vs 31.71±18.1, P = 0.013) (Fig3A). GC-based therapies (GC alone or in combination with IM/RTX) were used more frequently in hypocomplementemia group (100% vs 85.7%, P=0.001). Higher IgG4-RI and more organs involved in hypocomplementemia group, indicating that the different treatment regimens between the two groups.
To study the impact of hypocomplementemia on treatment outcomes, 45 patients with hypocomplementemia and 129 patients with normal complement followed-up more than 24 months were compared. The results showed that although IgG4-RD RI score was higher in hypocomplementemia group at baseline and 6th month follow-up(P<0.001 and P=0.005, respectively), there was no significant difference between two groups at 12, 24, 36 and 48 months after therapy(Fig. 3C). Relapse rate was observed in 20.0%, 26.7%, 33.0%, 33.0% and 49.0% of the patients with hypocomplementemia at 12, 24, 36, 48 and 60 months after therapy, respectively. There was no significant difference of relapse rate in two groups (P=0.7559)(Fig.3D).