Clinical Relevance of Serum Galactose Deficient IgA1 in Patients with IgA Nephropathy

New biomarkers of IgA nephropathy (IgAN) are needed for non-invasive diagnosis and appropriate treatment. There is emerging evidence that galactose deficient IgA1 (Gd-IgA1) is a pivotal molecule in the pathogenesis of IgAN. However, few studies have investigated the role of Gd-IgA1 as a biomarker in IgAN. In this study, we investigated the clinical relevance of serum Gd-IgA1 levels in patients with IgAN. Two hundred and thirty biopsy-proven IgAN patients, 74 disease controls (patients with non-IgAN nephropathy), and 15 healthy controls were enrolled in this study. Levels of serum Gd-IgA1 were measured using an ELISA kit in serum samples obtained the day of renal biopsy. We compared levels of serum Gd-IgA1 according to the type of glomerular disease and analyzed the association between Gd-IgA1 levels and clinical and pathological parameters in patients with IgAN. We then divided IgAN patients into two groups according to Gd-IgA1 level and investigated the predictive value of Gd-IgA1 for progression of chronic kidney disease (CKD). Serum Gd-IgA1 levels were significantly higher in IgAN patients than disease controls and healthy controls. In patients with IgAN, serum Gd-IA1 levels were significantly correlated with estimated glomerular filtration rate, serum IgA level, and tubular atrophy/interstitial fibrosis. CKD progression was more frequent in IgAN patients with higher serum Gd-IgA1 levels than in those with lower serum Gd-IgA1 levels. Cox proportional hazard models showed that high GdIgA1 level was an independent risk factor for CKD progression after adjusting for several confounders. Our results suggest that serum Gd-IgA1 level is a useful diagnostic and prognostic marker in IgAN patients. Further studies with a larger sample size and longer follow-up duration are needed.


Introduction
IgA nephropathy (IgAN) is the most frequent form of primary glomerulonephritis and one of the important causes of chronic kidney disease (CKD) worldwide [1]. The clinical course and disease prognosis of IgAN patients vary, and about 20-40% of patients reach end stage renal disease (ESRD) within 20 years of diagnosis [2,3]. Therefore, early diagnosis, risk prediction for disease progression, and appropriate treatment are important in IgAN. However, the pathogenesis of this disease is not yet fully understood and curative treatment strategies remain to be established [4]. Although the current gold standard diagnostic and prognostic method for IgAN is renal biopsy, it is not frequently performed in the real clinical field due to some limitations and concerns about complications [4,5]. Thus, it is necessary to identify non-invasive biomarkers that can be used to diagnose IgAN and assess activity and outcomes of the disease.

Baseline Characteristics of Study Population
The baseline characteristics of the participants in the IgAN group, disease control group, and healthy control group are summarized in Table 1. The median age of the IgAN patients was 41.00 years, and 50% were male. IgAN patients showed significantly higher levels of serum albumin than patients with MN, MCD, and LN, and lower levels of serum albumin than patients with TBMD and healthy controls. Patients with MN and MCD excreted significantly more urinary protein than IgAN patients. Levels of serum Gd-IgA1 were significantly elevated in patients with IgAN compared to disease controls and healthy controls (Table 1 and Figure 1). Figure 2 shows the correlation between serum Gd-IgA1 level and clinical parameters in 230 IgAN patients. The Gd-IgA1 level showed a weak negative correlation with eGFR (r = −0.146, p = 0.026, Figure 2A) and a positive correlation with serum IgA level (r = 0.550, p < 0.001, Figure 2B). However, neither urine PCR nor albumin was significantly correlated with Gd-IgA1 level (p = 0.127 and p = 0.065, respectively) ( Figure 2C,D). Based on the Oxford classification, serum Gd-IgA1 level was significantly elevated in patients with tubular atrophy/interstitial fibrosis (T0, 9.15 (6.93-12.18) vs. T1-2, 10.93 (8.45-16.69), p = 0.024) ( Table 2).  Figure 2 shows the correlation between serum Gd-IgA1 level and clinical parameters in 230 IgAN patients. The Gd-IgA1 level showed a weak negative correlation with eGFR (r = −0.146, p = 0.026, Figure 2A) and a positive correlation with serum IgA level (r = 0.550, p < 0.001, Figure 2B). However, neither urine PCR nor albumin was significantly correlated with Gd-IgA1 level (p = 0.127 and p = 0.065, respectively) ( Figure 2C,D). Based on the Oxford classification, serum Gd-IgA1 level was significantly elevated in patients with tubular atrophy/interstitial fibrosis (T0, 9.15 (6.93-12.18) vs. T1-2, 10.93 (8.45-16.69), p = 0.024) ( Table 2).

Discussion
In this study, we measured serum Gd-IgA1 level and investigated its clinical relevance in patients with IgAN. Our major findings were (1) serum Gd-IgA1 levels in patients with IgAN were significantly higher than those in disease controls and healthy controls; (2) serum Gd-IgA1 level was negatively correlated with eGFR and positively correlated with serum IgA in patients with IgAN; (3) serum Gd-IgA1 level was significantly elevated in IgAN patients with tubular atrophy/interstitial fibrosis; (4) CKD progression was more frequent in IgAN patients with a higher level of serum Gd-IgA1 than those with a lower serum Gd-IgA1 level; (5) higher serum Gd-IgA1 was an independent predictor of CKD progression in patients with IgAN.
Although there have been remarkable advances since IgAN was first described by Berger et al. in 1968 [16], the pathogenesis of the disease is not yet fully understood and there are currently no disease-specific biomarkers that are reliable and useful in clinical practice [4,11]. Despite the proposal of several candidate biomarkers in recent years, these biomarkers lack sensitivity and specificity [5]. Pathologic findings and nonspecific clinical parameters such as eGFR, urine protein excretion, and blood pressure are therefore currently used to assess disease activity and prognosis in IgAN [17,18].
The multi-hit hypothesis of IgAN pathogenesis is widely accepted. This multi-hit hypothesis proposes the following disease pathogenesis: first, an increase in aberrant glycosylation of IgA1 leading to overproduction of Gd-IgA1; second, synthesis of antibodies that recognize Gd-IgA1; third, formation of pathogenic immune complexes; and fourth, mesangial deposition of these complexes and initiation of renal injury [4,9]. Several studies have provided evidence supporting the multi-hit hypothesis, and Gd-IgA1 is therefore drawing attention as a potential biomarker of IgAN [19].
Previous studies have revealed that IgAN patients have significantly higher levels of serum Gd-IgA1 than patients with non-IgAN glomerular diseases and healthy subjects [8,12]. Consistent with prior studies, we observed a significantly increased serum Gd-IgA1 level in IgAN patients compared with disease controls and healthy controls. Considering that Gd-IgA1 level could be affected by serum IgA level, we further analyzed the serum Gd-IgA1 to IgA ratio (Supplementary Table S1). Serum Gd-IgA1 to IgA ratio was significantly elevated in patients with IgAN as compared with in patients with MN, MCD, and LN, and healthy controls. TBMD patients had lower Gd-IgA1 to IgA ratio than IgAN patients, but there was no statistical significance. Further studies are needed to validate our findings and confirm the clinical relevance of the Gd-IgA1 to IgA ratio.
Several study groups have demonstrated that Gd-IgA1 has clinical significance in patients with IgAN. Zhao et al. [20] showed that elevated serum Gd-IgA1 levels were associated with aggravation of urinary protein excretion and increased risk of renal function deterioration in IgAN patients. Other studies have also demonstrated that serum Gd-IgA1 levels were correlated with disease severity and renal outcome of IgAN [11,12]. In our study, serum level of Gd-IgA1 was correlated with eGFR, and the frequency of CKD progression was greater in IgAN patients with a higher serum Gd-IgA1 level than those with a lower serum Gd-IgA1 level. Multivariate Cox regression analysis revealed that higher serum Gd-IgA1 level was an independent risk factor for CKD progression.
Associations between serum Gd-IgA1 level and pathologic findings have also been reported. Xu et al. [21] showed that serum Gd-IgA1 level was closely associated with pathologic phenotypes in IgAN. They compared serum IgA1 glycosylation between IgAN patients with different pathologic phenotypes and observed that levels of α2,6 sialic acid and galactose of serum IgA1 were significantly lower in patients with focal proliferative sclerosing IgAN than those with in mild mesangial proliferative IgAN. Wada et al. [11] reported that serum Gd-IgA1 levels were significantly higher in IgAN patients with segmental sclerosis and tubular atrophy/interstitial fibrosis than those who did not have this condition. Consistent with this study, we also observed that serum Gd-IgA1 level was associated with tubular atrophy/interstitial fibrosis in IgAN patients. To our knowledge, the mechanism for the association between serum Gd-IgA1 levels and tubulointerstitial lesion has not been studied previously. Zhang et al. [10] showed the relationship between serum Gd-IgA1 levels and deposition of mesangial Gd-IgA1. Mesangial cells injured by deposition of Gd-IgA1 are reported to promote glomerulotubular cross-talk by secreting mediators such as cytokines and complements [22,23]. We plan to conduct further study assuming that this process might be related to the association between serum Gd-IgA1 level and tubular atrophy/interstitial fibrosis in IgAN patients. Further studies are needed to elucidate the underlying mechanisms.
A snail helix aspersa agglutinin (HAA) lectin-based assay is currently used to measure serum Gd-IgA1 level. Although the HAA lectin-based assay is a useful research tool and is widely used, the assay is complex to perform and there are issues with the bioactivity and stability of purified lectin [10,12]. Recently, a Gd-IgA1-specific antibody named KM55 was developed for lectin-independent assays [24]. Suzuki et al. demonstrated that KM55 recognizes Gd-IgA1 in IgAN patients as well as the HAA lectin-based assay does [24,25], and some studies have measured Gd-IgA1 using this assay and investigated the clinical significance of Gd-IgA1 in patients with IgAN [10,12]. In this study, we measured serum Gd-IgA1 levels using a lectin-independent assay and observed the similar trends to those reported in previous studies.
Our study had some potential limitations. First, the level of serum Gd-IgA1 was measured only once, which may have resulted in an incorrect classification of patients. To overcome this limitation, we are planning to build an additional independent cohort and test the diagnostic value of Gd-IgA1 as a biomarker. Using the cohort, we also plan to monitor Gd-IgA1 levels during the follow-up period and determine the prognostic value of Gd-IgA1 and its ability to assess the therapeutic effect. Second, urine protein level was measured in a spot urine sample. Third, the cut-off value of higher and lower serum Gd-IgA1 was suggested with appropriate calculation, but further studies with large sample sizes are needed to ascertain the reliability of the cut-off value. Despite these limitations, we demonstrated the clinical relevance of Gd-IgA1 by performing a long-term follow-up (mean follow-up period over 3 years) in a relatively large number of IgAN patients.
In summary, we found that serum Gd-IgA1 levels were noticeably elevated in patients with IgAN and were significant associated with clinicopathological variables. Higher serum Gd-IgA1 level predicted CKD progression in IgAN patients. Serum Gd-IgA1 is therefore a potential disease-specific biomarker for diagnosis and assessment of the disease progression of IgAN. Further studies with a larger sample size and longer follow-up are needed to confirm our findings.