Skip to main content
NCBI home page
Phenomics logoLink to Phenomics
. 2024 Mar 21;4(2):146–157. doi: 10.1007/s43657-023-00138-6

Clinical Application of Polygenic Risk Score in IgA Nephropathy

Linlin Xu 1,2,3,4, Ting Gan 1,2,3,4, Pei Chen 1,2,3,4, Yang Liu 1,2,3,4, Shu Qu 1,2,3,4, Sufang Shi 1,2,3,4, Lijun Liu 1,2,3,4, Xujie Zhou 1,2,3,4,✉, Jicheng Lv 1,2,3,4, Hong Zhang 1,2,3,4
PMCID: PMC11169313  PMID: 38884057

Abstract

Genome-wide association studies (GWASs) have identified 30 independent genetic variants associated with IgA nephropathy (IgAN). A genetic risk score (GRS) represents the number of risk alleles carried and thus captures an individual's genetic risk. However, whether and which polygenic risk score crucial for the evaluation of any potential personal or clinical utility on risk and prognosis are still obscure. We constructed different GRS models based on different sets of variants, which were top single nucleotide polymorphisms (SNPs) reported in the previous GWASs. The case–control GRS analysis included 3365 IgAN patients and 8842 healthy individuals. The association between GRS and clinical variability, including age at diagnosis, clinical parameters, Oxford pathology classification, and kidney prognosis was further evaluated in a prospective cohort of 1747 patients. Three GRS models (15 SNPs, 21 SNPs, and 55 SNPs) were constructed after quality control. The patients with the top 20% GRS had 2.42—(15 SNPs, p = 8.12 × 10–40), 3.89—(21 SNPs, p = 3.40 × 10–80) and 3.73—(55 SNPs, p = 6.86 × 10–81) fold of risk to develop IgAN compared to the patients with the bottom 20% GRS, with area under the receiver operating characteristic curve (AUC) of 0.59, 0.63, and 0.63 in group discriminations, respectively. A positive correlation between GRS and microhematuria, mesangial hypercellularity, segmental glomerulosclerosis and a negative correlation on the age at diagnosis, body mass index (BMI), mean arterial pressure (MAP), serum C3, triglycerides can be observed. Patients with the top 20% GRS also showed a higher risk of worse prognosis for all three models (1.36, 1.42, and 1.36 fold of risk) compared to the remaining 80%, whereas 21 SNPs model seemed to show a slightly better fit in prediction. Collectively, a higher burden of risk variants is associated with earlier disease onset and a higher risk of a worse prognosis. This may be informational in translating knowledge on IgAN genetics into disease risk prediction and patient stratification.

Supplementary Information

The online version contains supplementary material available at 10.1007/s43657-023-00138-6.

Keywords: Genomics, IgA nephropathy, Polygenic score, Prognosis, Risk prediction

Introduction

Immunoglobulin A (IgA) nephropathy (IgAN), the most common form of glomerulonephritis, is one of the leading causes of end-stage kidney disease (ESKD) among Asian populations (D'Amico 1987; Floege and Amann 2016Tsukamoto et al. 2009). The pathogenesis of the disease remains unclear, but its development is driven by both genetic susceptibility and environmental risk factors (Barsoum 2010; Kiryluk et al. 2014; Li and Yu 2018; Neugut and Kiryluk 2018; Wakai et al. 2002). The prevalence of IgAN shows striking ethnic variation, being more prevalent in East Asian ancestry and less prevalent in African ancestry compared with Europeans (Kiryluk et al. 2012). Although not generally considered a Mendelian hereditary disease, marked interethnic differences (Schena 1990) and familial clustering (Julian et al. 1985; Scolari et al. 1999; Wyatt et al. 1987), along with subclinical renal abnormalities among relatives of IgAN cases (Lai et al. 2016), have suggested a strong genetic component. Genetic factors undoubtedly make a substantial contribution to the occurrence and development of the disease, with an estimated heritability of 40–50% (Hastings et al. 2010; Kiryluk et al. 2011). In the last 15 years, genome-wide association studies (GWASs) have successfully identified genetic loci associated with IgAN risk (Feehally et al. 2010; Gharavi et al. 2011; Kiryluk et al. 2014, 2023; Li et al. 2015, 2020; Yu et al. 2011). In the more recent GWAS meta-analysis, Kiryluk et al. (2023) identified 30 independent genome-wide significant risk loci jointly explaining 11% of disease risk, which also suggested IgAN was a complex disease with polygenic involvement. However, individual genetic loci have a weak effect on complex traits/diseases. Hence, it has been extensively suggested that it is of great significance to add the effects of these genetic loci to a genetic risk score (GRS), which can be used to conduct disease risk stratification or predict prognostic outcomes and response to therapy to facilitate the realization of precision medicine (Lambert et al. 2019; Natarajan et al. 2017; Torkamani et al. 2018). For example, genetic risk scores have been validated to exert clinical utility in coronary heart disease and breast cancer (Abraham et al. 2016; Elliott et al. 2020; Hurson et al. 2022; Inouye et al. 2018; Kullo et al. 2016; Lakeman et al. 2019; Mavaddat et al. 2019; Zhang et al. 2018). A key advantage of a GRS is that it can be quantifiable at the time of birth, a long time before the onset of traditional clinical risk factors, which can be used to predict specific disease risks early in life.

There are also a few numbers of studies on genetic risk scores for IgAN. A five-single nucleotide polymorphism (SNP) GRS showed that the disease risk varied up to tenfold between individuals with no protective alleles and those with five or more (Gharavi et al. 2011). A standardized genetic risk based on 15 SNPs was associated with age at disease onset but was not associated with progression (Kiryluk et al. 2014). Our earlier research found that unweighted GRS by nine SNPs (uwGRS9) ≥ 16 was an independent predictor for ESKD in IgAN, with a relative risk of 2.52 (Zhou et al. 2014). Recently, a genome-wide polygenic risk score (GPS) of IgAN has been shown to be associated with early disease onset, the increased lifetime risk of kidney failure, as well as hematuria in the independent German chronic kidney disease (GCKD) Study (Kiryluk et al. 2023). The majority of these studies are based on European populations. More widespread replications in multiple populations are needed. In addition, early reports were limited in sample sizes and numbers of susceptibility loci, which may be inadequate in evaluating the significance of genetic risk score in predicting prognosis.

To shift from studying the genetics of susceptibility to precision medicine and explore the clinical utility of this genetic information, we not only evaluated the potential of GRS as a meaningful index of risk to stratify IgAN risk in the general population but also appraised its efficacy in characterizing IgAN-related phenotypes and assessed the critical time dimension to provide information about individual progression in our large cohort with complete genetic profiles and in-depth clinical data.

Materials and Methods

Study Population

The study sample consisted of five datasets after quality controls: the first dataset comprised 1171 patients and 891 healthy individuals genotyped with the Illumina 610-quad BeadChip, the second dataset comprised 485 patients and 3866 healthy individuals genotyped with the Illumina Infinium OmniZhongHua-8 v1.3 array, the third dataset comprised 1110 patients and 1240 healthy individuals genotyped with Infinium Global Screening Array-24 v1.0 BeadChip, the fourth dataset comprised 312 patients and 1545 healthy individuals genotyped with Infinium Global Screening Array-24 v2.0 BeadChip, and the fifth dataset comprised 417 patients and 1559 healthy individuals genotyped with the Infinium Chinese Genotyping Array-24 v1.0 BeadChip. The details for the recruitment of patients, genotyping, and genotype quality control were previously reported (Kiryluk et al. 2023). Compared to previous reports, the fourth dataset and the fifth dataset were novel enrolled cohorts.

In total, 3495 patients and 9101 healthy controls recruited in the renal division of Peking University First Hospital were used to perform the association analysis between a single SNP selected and susceptibility to IgAN. Among these patients, 1803 patients were regularly followed (PKU-IgAN Cohort) from 1997 to 2020. All IgAN patients were diagnosed by biopsy and those with less than eight glomeruli per biopsy section were excluded. Patients with IgAN secondary to Henoch-Schönlein purpura, lupus nephritis, chronic liver diseases, and other immunologic disorders were excluded. The controls were unrelated to the cases, derived from the same populations as the cases, and had no known kidney disease.

Detailed baseline data were collected from patients at the time of kidney biopsy. Kidney biopsies were scored according to established criteria for the Oxford MEST-C scoring system (mesangial hypercellularity (M); endocapillary hypercellularity (E); segmental glomerulosclerosis (S); tubular atrophy/interstitial fibrosis (T); crescents (C)) (Cattran et al. 2009; Roberts et al. 2009; Trimarchi et al. 2017), which were reassessed by the same pathologist. Time-averaged (TA) proteinuria, serum albumin, mean arterial pressure (MAP), and microhematuria were also calculated (Li et al. 2014). In this study, the primary outcome was ESKD, and the secondary outcome (combined event) was the combination of ESKD or ≥ 50% reduction in estimated glomerular filtration rate (eGFR) after diagnostic kidney biopsy. ESKD was defined as eGFR less than 15 ml/min/1.73 m2, dialysis, or kidney transplantation.

This study was approved by the ethics committee of Peking University First Hospital [IRB number 2020Y197] and was conducted in accordance with the principle of the Helsinki Declaration. Written informed consent was provided by all participants.

SNPs Selection for GRS

Genome-wide significant loci (p < 5.0E−08) associated with IgAN based on the previous GWASs (Kiryluk et al. 2012, 2014, 2023; Li et al. 2015, 2020; Yu et al. 2011) and suggestive loci in the combined trans-ethnic meta-analysis or additionally East Asians-specific suggestive loci (p < 5.0E−05) from the more recent GWAS-meta study by Kiryluk et al. (2023) were selected. Based on the above criteria, a total of 116 SNPs were retrieved. In this study, a total of 107 variants remained for subsequent analyses, excluding nine variants being not genotyped or imputed in our data. Details of the 107 SNPs were shown in Table S1.

GRS Construction

We constructed 15 SNPs-GRS suggested by Kiryluk et al. in 2014, 30 SNPs-GRS based on more recently reported genome-wide significant lead SNPs (Kiryluk et al. 2023), and 107 SNPs-GRS including all genome-wide significant and suggestive loci, respectively. To ensure the accuracy of GRS calculation, SNPs and samples with a call rate < 0.95 were removed. To maximize the effectiveness, a corresponding proxy SNP with r2 > 0.8 in linkage disequilibrium (LD) was adopted when the lead SNP was missing (Natarajan et al. 2017). Additionally, only significantly associated SNPs (p < 0.05) in our datasets were adopted in further model constructions. When several lead SNPs were reported in a single locus, only independent variants were selected (LD r2 < 0.5), in case of double counting of the same gene (Fig. 1). Finally, a total of 3365 patients and 8842 healthy controls as well 55 SNPs passed quality control, which was left to construct 15 SNPs-GRS, 21 SNPs-GRS, and 55 SNPs-GRS according to the above study design. The detailed process is illustrated in Fig. 1. The GRS was the weighted sum of risk allele counts (zero, one, or two) multiplied by the natural log of the adjusted odds ratio (OR) for each of the individual loci (Dai et al. 2019). In this study, the effect sizes were directly obtained from the original GWAS reports. If a given variant was associated with IgAN susceptibility at GWAS significance in more than one GWAS, we selected the effect size with the minimal p value (Lu et al. 2022). The GRS was calculated by PLINK v1.9 (Purcell et al. 2007). GRS was standardized using the mean and variance of the control group for all subsequent analysis (Toulopoulou et al. 2019).

Fig. 1.

Fig. 1

Open in a new tab

The flow chart of this study. A total of 116 SNPs were retrieved based on previous GWASs (Kiryluk et al. 2012, 2014, 2023; Li et al. 2015, 2020; Yu et al. 2011), including genome-wide significant loci (p < 5.0E−08) associated with IgAN and suggestive loci in the combined trans-ethnic meta-analysis or additionally East Asians-specific suggestive loci (p < 5.0E−05) from the more recent GWAS-meta study by Kiryluk et al. (2023). Nine of these SNPs were not genotyped or imputed in our data. Firstly, we analyzed associations between disease susceptibility and 107 SNPs in 3495 patients and 9101 healthy controls (IgAN susceptibility association analysis). 1803 of the 3495 patients were routinely followed up, so we further analyzed the association between the 107 SNPs and the clinical phenotypes (sub-phenotype and prognosis analysis) in 1803 patients. After quality control, a total of 3365 patients and 8842 healthy controls with complete information of 55 SNPs were left to construct GRS models. To check the cumulative effect of these SNPs together, lastly, we also analyzed the association of GRS models with disease susceptibility, clinical subphenotypes, and disease prognosis

To enhance the robustness of the results, we further constructed 15 SNPs-uwGRS (unweighted GRS, uwGRS), 15 SNPs-standardized genetic risk model suggested originally (Kiryluk et al. 2014), 15 SNPs-wGRS (weighted GRS, wGRS) calculated by Plink, and 15 SNPs-wGRS calculated by PRSice (Choi and O'Reilly 2019), respectively. The uwGRS was the simple count of the total number of risk alleles rather than weighting by the effect of each SNP (Robson et al. 2017; Zhou et al. 2014). The correlation analysis between GRS models constructed by the above four methods was completed by linear correlation.

Statistical Analysis

We combined single SNP association results across different batches via a standard error-weighted meta-analysis using METAL (Willer et al. 2010). Covariates of the first two principal components were included for adjustment as previously reported (Kiryluk et al. 2014, 2023).

Continuous variables were summarized as mean ± standard deviation (SD) (normally distributed) or median (Interquartile Range) (non-normally distributed). Categorical variables were presented as frequencies (percent). The association of each allele with the risk of IgAN as well as clinical phenotypes were analyzed in an additive model using PLINK v1.9 (Purcell et al. 2007).

The difference in GRS between patients and controls was tested using the Student’s test. In addition, we also divided the participants into quintiles based on the GRS of controls. Logistic regression was used to assess the risk difference among quintiles. The performance of GRS was assessed by the area under the receiver operating characteristic curve (AUC). For association analysis of GRS and clinical characteristics, the correlation analysis was used for continuous variables, and linear regression was used for categorical variables. Multiple comparisons were corrected using the Bonferroni method. Renal survival time from the outcome event was calculated from the biopsy to the last follow-up. A Kaplan–Meier cumulative renal survival analysis was used to test the association between GRS and age at poor prognosis. The Cox proportional hazards regression model was used to evaluate the lifetime risk of incident ESKD/combined event with age as the time scale with the adjustments of sex and the first two genetic principal components, which referred to similar studies on polygenic risk score (Kiryluk et al. 2023; Lu et al. 2022; Northcutt et al. 2021). Patients lost to follow-up were censored and not included as events. Statistical analyses were performed by SPSS version 26.0 and R 4.2.0. Two-tailed p < 0.05 was considered statistically significant.

Results

Association Between Single SNPs and Susceptibility to IgAN

For the 107 SNPs available, 81 showed nominal significance (p < 0.05) with IgAN risk that was in the same direction as previously reported, whereas 39 variants showed significant associations even at a Bonferroni-corrected threshold (p value ranged from 4.48E−04 to 8.47E−26 < 0.05/107). The specific results are provided in Table S2.

Single SNP-Subphenotypes Correlation Analysis

The demographic characteristics, laboratory and histologic findings, and follow-up information of 1747 patients are shown in Table S3. The individual association between the susceptibility SNPs with characteristics was assessed, which included age at biopsy, the history of gross hematuria, eGFR, proteinuria, microhematuria, and serum IgA. The results of the association between 30 genome-wide significant SNPs (Kiryluk et al. 2023) and characteristics are shown in Table S4. We found that the risk allele A of rs3803800 (p = 0.02) and the risk allele G of rs2412971 (p = 0.03) were associated with an increased serum IgA level, which was concordant with the previous studies (Kiryluk et al. 2014; Yang et al. 2012; Yu et al. 2011; Zhou et al. 2014). For rs3803800 (TNF Superfamily Member 13, TNFSF13), the serum IgA concentrations (g/L, mean ± SD) were 3.20 ± 1.20, 3.35 ± 1.17, and 3.36 ± 1.26 for rs3803800 GG, AG, and AA, respectively (Fig. S1a). For rs2412971 (HORMA Domain Containing 2, HORMAD2), the serum IgA concentrations were 3.20 ± 1.22, 3.22 ± 1.19, and 3.36 ± 1.20 for rs2412971 AA, AG, and GG, respectively (Fig. S1b).

We also assessed the association between SNPs and prognosis. Of the 107 SNPs, the risk alleles of five SNPs were associated with ESKD, including rs12716641 (Defensin Alpha, DEFA) (HR 1.34, 95% CI 1.01–1.77, p = 0.04), rs75413466 (LYN Proto-Oncogene, Scr Family Tyrosine Kinase, LYN) (HR 1.55, 95% CI 1.06–2.27, p = 0.02), rs10088534 (Outer Dense Fiber Of Sperm Tails 1, ODF1) (HR 0.76, 95% CI 0.62–0.94, p = 0.01), rs2033562 (Outer Dense Fiber Of Sperm Tails 1/KLF Transcription Factor 10, ODF1/KLF10) (HR 0.76, 95% CI 0.62–0.93, p = 0.01), and rs79246816 (Kelch Like Family Member 1, KLHL1) (HR 1.95, 95% CI 1.18–3,24, p = 0.01). The risk alleles of nine SNPs were associated with ESKD or ≥ 50% reduction in eGFR, rs4957300 (Prostaglandin E Receptor 4, PTGER4) (HR 0.82, 95% CI 0.69–0.99, p = 0.04), rs2738048 (DEFA) (HR 1.28, 95% CI 1.05–1.56, p = 0.02), rs12716641 (DEFA) (HR 1.42, 95% CI 1.12–1.79, p = 4.00 × 10–3), rs75413466 (LYN) (HR 1.38, 95% CI 1.00–1.91, p = 0.05), rs10088534 (ODF1) (HR 0.81, 95% CI 0.69–0.96, p = 0.01), rs2033562 (ODF1/KLF10) (HR 0.81, 95% CI 0.69–0.96, p = 0.02), rs4077515 (Caspase Recruitment Domain Family Member 9, CARD9) (HR 1.23, 95% CI 1.03–1.46, p = 0.02), rs4227 (Mannose-P-Dolichol Utilization Defect 1, MPDU1) (HR 1.23, 95% CI 1.01–1.49, p = 0.04) and rs7227084 (Myelin Basic Protein, MBP) (HR 1.30, 95% CI 1.01–1.67, p = 0.04). The correlation analysis results of 107 SNPs with clinical phenotypes and prognosis are shown in Table S5.

The Construction of GRS and Distribution in IgAN Cases and Controls

In GRS construction, a total of 55 SNPs passed SNP quality control, and 3365 patients and 8842 controls passed sample quality control, which was further used to construct 15 SNPs-GRS, 21 SNPs-GRS, and 55 SNPs-GRS according to the study design, respectively (Fig. 1). For 15 SNPs-GRS, we further constructed 15 SNPs-uwGRS, 15 SNPs-standardized genetic risk model, 15 SNPs-wGRS based on PLINK software, and 15 SNPs-wGRS based on PRSice software according to the four methods mentioned above. The correlation analysis results of 15 SNP-GRSs constructed by these four methods are shown in Fig. S2. The correlation coefficient was 1.00 between 15 SNPs-wGRS calculated by Plink and 15 SNPs-wGRS calculated by PRSice. The correlation coefficient was 0.95 between 15 SNPs-uwGRS and 15 SNPs-wGRS. Therefore, in the following study, we constructed 15 SNPs-GRS, 21 SNPs-GRS, and 55 SNPs-GRS using the weighted method according to the study design.

The distribution of 15 SNPs-GRS (0.00 ± 1.00 vs. 0.29 ± 0.97; p = 1.40 × 10–47), 21 SNPs-GRS (0.00 ± 1.00 vs. 0.46 ± 0.97; p = 8.34 × 10–115) and 55 SNPs-GRS (0.00 ± 1.00 vs. 0.47 ± 1.00; p = 2.99 × 10–115) between patients and healthy controls were significantly different (Fig. 2). The GRS was higher in patients compared to the one in controls.

Fig. 2.

Fig. 2

Open in a new tab

The different GRS distribution and performance according to Receiver-operating curve (ROC). Density plots showing the distribution of standardized GRS among IgAN patients and healthy controls for a 15 SNPs-GRS, b 21 SNPs-GRS, and c 55 SNPs-GRS, respectively. The ROC curve of d15 SNPs-GRS, e 21 SNPs-GRS, and f 55 SNPs-GRS, respectively

The 21 SNPs-GRS (OR 1.62, 95% CI 1.55–1.69, p = 1.59 × 10–107) and 55 SNPs-GRS (OR 1.62, 95% CI 1.55–1.69, p = 1.89 × 10–108) had the same association with IgAN risk, which had a greater association with IgAN risk than 15 SNPs-GRS (OR 1.36, 95% CI 1.30–1.42, p = 1.30 × 10–45). We also observed a more marked gradient of IgAN risk across quintiles of 21 SNPs-GRS than 15 SNPs-GRS. The 21 SNPs-GRS had a greater OR of 3.89 (95% CI 3.38–4.47, p = 3.40 × 10–80) in the top quintile vs. the bottom quintile than the 15 SNPs-GRS (OR 2.42, 95% CI 2.13–2.77, p = 8.12 × 10–40). The 55 SNPs-GRS had a similar OR of 3.73 (95% CI 3.26–4.27, p = 6.86 × 10–81) in the top quintile vs. the bottom quintile compared to 21 SNPs-GRS (Table 1). The performance of GRS to discriminate IgAN patients from controls was 0.59 (p = 2.42 × 10–50), 0.63 (p = 3.54 × 10–113), and 0.63 (p = 6.07 × 10–118) (AUC) for 15 SNPs-GRS, 21 SNPs-GRS, and 55 SNPs-GRS, respectively (Fig. 2).

Table 1.

Disease of susceptibility to IgAN based on quintiles of GRS

GRS AUC Group Control IgAN OR 95% CI p
Mean GRS n (%) Mean GRS n (%)
15 SNPs-GRS 0.59 Q1 − 1.51 1768 (20%) − 1.53 420 (13%) 1.00
Q2 − 0.51 1771 (20%) − 0.51 546 (16%) 1.30 1.13–1.50 3.62 × 10–4
Q3 0.13 1766 (20%) 0.15 651 (19%) 1.55 1.35–1.78 6.26 × 10–10
Q4 0.64 1771 (20%) 0.66 731 (22%) 1.74 1.52–1.99 2.59 × 10–15
Q5 1.25 1766 (20%) 1.30 1017 (30%) 2.42 2.13–2.77 8.12 × 10–40
21 SNPs-GRS 0.63 Q1 − 1.45 1768 (20%) − 1.40 315 (10%) 1.00
Q2 − 0.51 1770 (20%) − 0.50 480 (14%) 1.52 1.30–1.78 1.47 × 10–7
Q3 0.05 1771 (20%) 0.06 593 (18%) 1.88 1.62–2.19 3.60 × 10–16
Q4 0.57 1765 (20%) 0.59 753 (22%) 2.40 2.07–2.77 2.82 × 10–31
Q5 1.35 1768 (20%) 1.44 1224 (36%) 3.89 3.38–4.47 3.40 × 10–80
55 SNPs-GRS 0.63 Q1 − 1.46 1772 (20%) − 1.46 348 (10%) 1.00
Q2 − 0.50 1767 (20%) − 0.49 427 (13%) 1.23 1.05–1.44 9.00 × 10–3
Q3 0.05 1772 (20%) 0.06 596 (18%) 1.71 1.48–1.99 9.40 × 10–13
Q4 0.56 1763 (20%) 0.58 699 (21%) 2.02 1.75–2.33 1.59 × 10–21
Q5 1.35 1768 (20%) 1.43 1295 (38%) 3.73 3.26–4.27 6.86 × 10–81
Open in a new tab

GRS genetic risk score, SNP single nucleotide polymorphisms, AUC area under curve, IgAN immunoglobulin A (IgA) nephropathy, OR odds ratio, 95% CI 95% confidential interval

Associations Between GRS and IgAN Subphenotypes

IgAN is the most common primary glomerular disease worldwide. The diagnostic histologic hallmark is dominant or codominant IgA staining on kidney biopsy (Wyatt and Julian 2013); however, patients may present with various clinical syndromes ranging from asymptomatic abnormalities noted on urinalysis to rapidly progressive glomerulonephritis (Lai et al. 2016). It has been widely accepted that the clinical characteristics may be intermediate phenotypes between genetic risk factors and the full-blown disease. It is of pathological significance to check the sub-phenotype genetic associations to further understand the underlying pathogenesis of IgAN. Individually, it has been confirmed that several risk alleles were associated with IgAN sub-phenotypes, such as associations between Myotubularin Related Protein 3 (MTMR3) variants and the level of serum IgA (Wang et al. 2023). It was also significant to check the associations between sub-phenotypes and cumulative risk alleles captured by GRS. Therefore, we performed a comprehensive genetic analysis, including demographic characteristics (age at diagnosis, body mass index (BMI), MAP, the history of precursor infection, and the history of gross hematuria), laboratory findings at the time of renal biopsy (serum creatinine, eGFR, serum albumin, serum uric acid, serum IgA level, serum C3 level, low-density lipoprotein, triacylglycerol, and total cholesterol, microhematuria, and 24 h-proteinuria), Oxford pathology classification and TA-MAP/microhematuria/proteinuria/albumin in follow-up.

For the demographic characteristics, the age at diagnosis was negatively associated with 55 SNPs-GRS (Pearson’s r = − 0.05, p = 0.04) and showed a marginal significance with 15 SNPs-GRS (Pearson’s r = − 0.04, p = 0.09) and 21 SNPs-GRS (Pearson’s r = − 0.04, p = 0.08). When the patients were grouped according to 15 SNPs-GRS quintiles, age at biopsy varied between groups (p = 0.02). Although there was no difference in age at biopsy among groups when patients were grouped by 21 SNPs-GRS/55 SNPs-GRS, there was a negative correlation trend (Fig. S3). The BMI and MAP were negatively associated with all three GRS models. For the baseline laboratory findings, the ln(microhematuria) and IgA/C3 were positively associated with 21 SNPs-GRS and 55 SNPs-GRS, and the ln(microhematuria) showed a marginal significance (Pearson’s r = 0.05, p = 0.06) with 15 SNPs-GRS. The serum C3 (Pearson’s r range from − 0.10 to − 0.06, p value range from < 0.001 to 0.01) was negatively associated with all three GRS models. The triglyceride levels were negatively associated with 21 SNPs-GRS (Spearman’s r = − 0.06, p = 0.01) and 55 SNPs-GRS (Spearman’s r = − 0.06, p = 0.02), which showed a marginal significance (Spearman’s r = − 0.05, p = 0.06) with 15 SNPs-GRS. For the Oxford MEST-C scores, the mesangial hypercellularity was positively associated with the 15 SNPs-GRS (beta = 0.03, p = 0.04), the 21 SNPs-GRS (beta = 0.03, p = 0.01), and the 55 SNPs-GRS (beta = 0.04, p = 0.003). And the segmental glomerulosclerosis was positively associated with the 21 SNPs-GRS (beta = 0.03, p = 0.01) and the 55 SNPs-GRS (beta = 0.03, p = 0.03), which showed a marginal significance with 15 SNPs-GRS (beta = 0.02, p = 0.06). For all the associated parameters, after Bonferroni correction (0.05/78 tests), the serum C3 and IgA/C3 were still associated with 21 SNPs-GRS. The specific results are shown in Tables 2, 3, and 4.

Table 2.

The association between GRS and clinical characteristics (normally distributed) of IgAN

Characteristics 15 SNPs-GRS 21 SNPs-GRS 55 SNPs-GRS
Pearson’s r p Pearson’s r p Pearson's r p
Age (n = 1747) − 0.04 0.09 − 0.04 0.08 − 0.05 0.04
BMI (n = 1695) − 0.08 1.00 × 10–3 − 0.07 3.00 × 10–3 − 0.08 1.00 × 10–3
eGFR (n = 1747) 0.01 0.74 − 0.05 0.06 − 0.02 0.41
MAP (n = 1745) − 0.05 0.03 − 0.05 0.03 − 0.05 0.02
Alb (n = 1729) 0.02 0.34 0.03 0.27 0.02 0.33
UA (n = 1727) − 0.02 0.44 2.00 × 10–3 0.93 3.00 × 10–3 0.89
ln(microhematuria) (n = 1588) 0.05 0.06 0.06 0.02 0.05 0.03
IgA (n = 1604) 0.01 0.72 0.07 0.01 0.04 0.16
C3 (n = 1559) − 0.06 0.01 − 0.10 7.00 × 10–5 − 0.08 1.00 × 10–3
IgA/C3 (n = 1548) 0.04 0.14 0.11 2.30 × 10–5 0.07 0.01
TA-MAP (n = 1662) − 0.04 0.15 3.00 × 10–3 0.90 − 0.01 0.63
ln(TA-microhematuria) (n = 1187) 0.07 0.02 0.08 4.00 × 10–3 0.07 0.02
Open in a new tab

Significant associations after Bonferroni correction highlighted in bold

GRS genetic risk score, SNP single nucleotide polymorphisms, BMI body mass index, eGFR estimated glomerular filtration rate, MAP mean arterial pressure, Alb albumin, UA uric acid, TA-MAP time average-mean arterial pressure

Table 3.

The association between GRS and clinical characteristics (non-normally distributed) of IgAN

Characteristics 15 SNPs-GRS 21 SNPs-GRS 55 SNPs-GRS
Spearman’s r p Spearman's r p Spearman’s r p
Scr (n = 1747) − 0.01 0.67 0.05 0.04 0.03 0.29
UTP (n = 1741) − 0.04 0.08 − 0.04 0.11 − 0.05 0.04
LDL (n = 1678) − 0.03 0.26 − 0.01 0.63 − 0.02 0.42
TG (n = 1696) − 0.05 0.06 − 0.06 0.01 − 0.06 0.02
TCHO (n = 1697) − 0.03 0.18 − 0.04 0.12 − 0.03 0.19
TA-P (n = 1654) − 4.00 × 10–3 0.88 0.04 0.13 0.02 0.48
TA-Alb (n = 1607) − 0.05 0.06 − 0.07 0.01 − 0.06 0.03
Open in a new tab

GRS genetic risk score, SNP single nucleotide polymorphisms, Scr serum creatinine, UTP urine protein, LDL low density lipoprotein, TG triglycerides, TCHO total cholesterol, TA-P time average-proteinuria, TA-Alb time average-albumin

Table 4.

The association between GRS and clinical characteristics (categorical variables) of IgAN

Characteristics 15 SNPs-GRS 21 SNPs-GRS 55 SNPs-GRS
Beta p Beta p Beta p
The history of infection (n = 1743) 2.34 × 10–4 0.98 0.02 0.18 4.00 × 10–3 0.74
The history of gross hematuria (n = 1743) 0.04 0.12 0.01 0.21 0.01 0.32
M (n = 1675) 0.03 0.04 0.03 0.01 0.04 3.00 × 10–3
E (n = 1675) − 2.00 × 10–3 0.88 − 8.00 × 10–3 0.48 − 3.00 × 10–3 0.79
S (n = 1675) 0.02 0.06 0.03 0.01 0.03 0.03
T (n = 1675) − 8.00 × 10–3 0.63 6.00 × 10–3 0.72 − 1.00 × 10–3 0.97
C (n = 1675) 0.02 0.25 0.02 0.16 0.03 0.12
Open in a new tab

GRS genetic risk score, SNP single nucleotide polymorphisms, M mesangial hypercellularity, E endocapillary hypercellularity, S segmental glomerulosclerosis, T tubular atrophy/interstitial fibrosis, C cellular/fibrocellular crescents

Association Between GRS and Prognosis of IgAN

During a median follow-up of 53 months, 192 patients (11%) had reached the endpoint of ESKD and 278 patients (16%) had reached the combined event (ESKD or ≥ 50% reduction in eGFR after diagnostic kidney biopsy).

In follow-up, the ln(TA-microhematuria) was positively associated with all three GRS models (Pearson’s r range from 0.07 to 0.08, p value range from 0.004 to 0.02). The TA-albumin was negatively associated with 21 SNPs-GRS (Spearman’s r = − 0.07, p = 0.01) and 55 SNPs-GRS (Spearman’s r = − 0.06, p = 0.03), which showed a marginal significance (Spearman’s r = − 0.05, p = 0.06) with 15 SNPs-GRS.

For the primary outcome, by univariate Cox regression analysis, the 15 SNPs-GRS was marginally associated with the lifetime risk of ESKD (HR 1.36, 95% CI 0.97–1.90, p = 0.07), and the 21 SNPs-GRS (HR 1.27, 95% CI 0.91–1.76, p = 0.16) and 55 SNPs-GRS (HR 1.25, 95% CI 0.89–1.76, p = 0.20) didn’t show the association with the lifetime risk of ESKD (Fig. 3a–c). After adjusting for sex, the first two genetic principal components, 15 SNPs-GRS (HR 1.33; 95% CI 0.95–1.86, p = 0.10), 21 SNPs-GRS (HR 1.20; 95% CI 0.86–1.68, p = 0.28), and 55 SNPs-GRS (HR 1.24, 95% CI 0.88–1.74, p = 0.22) were not associated with the lifetime risk of ESKD.

Fig. 3.

Fig. 3

Open in a new tab

Prognostic association of the GRS for IgA nephropathy. Survival analysis of a lifetime of ESKD for IgAN cases in the top 20% of a 15 SNPs-GRS distribution, b 21 SNPs-GRS distribution, and c 55 SNPs-GRS distribution, respectively. Survival analysis of a lifetime of a combined event (ESKD or ≥ 50% reduction in eGFR after diagnostic kidney biopsy) for IgAN cases in the top 20% of d 15 SNPs-GRS distribution, e 21 SNPs-GRS distribution, and f 55 SNPs-GRS distribution, respectively. The x-axis shows age, the y-axis shows survival probability without ESKD/combined event with the number of participants at risk at each age cut-off of 20, 40, 60 and 80 years depicted below

For the secondary outcome (a combined event), by univariate Cox regression analysis, individuals in the top 20% tail of the 15 SNPs-GRS distribution had 36% increased risk of a combined event compared to the remaining 80% (HR 1.36, 95% CI 1.03–1.80, p = 0.03), while individuals in the top 20% tail of the 21 SNPs-GRS distribution had 42% increased risk of a combined event compared to the remaining 80% (HR 1.42, 95% CI 1.09–1.86, p = 0.01). However, the risk of individuals in the top 20% tail of the 55 SNPs-GRS compared to the remaining 80% (HR 1.36, 95% CI 1.03–1.80, p = 0.03) was not higher than 21 SNPs-GRS, which was the same as 15 SNPs-GRS (Fig. 3d–f). After adjusting for sex, the first two genetic principal components, 15 SNPs-GRS (HR 1.35, 95% CI 1.02–1.78, p = 0.04), 21 SNPs-GRS (HR 1.37; 95% CI 1.05–1.80, p = 0.02), and 55 SNPs-GRS (HR 1.34; 95% CI 1.02–1.77, p = 0.04) were still associated with the progression of the disease.

Discussion

We tested the clinical utility of GRS-based established SNPs associated with IgAN in a large cohort of Chinese patients. We first replicated the majority of these loci, suggesting their real genetic effect. To determine their cumulative effect, we constructed and evaluated genetic risk scores based on different portfolios of different SNPs for IgAN and tested their potential for disease prediction, and further probed the relationship between three genetic risk scores, clinical characteristics, and prognosis. Compared with single SNP, GRS models were more significantly associated with disease susceptibility. All GRS models were associated with IgAN, in which the prediction power increased with increasing numbers of genome-wide significant SNPs selected. However, when suggestive SNPs were added, the prediction power didn’t increase significantly. This suggested that GRS based on genome-wide significant loci (30 or 21 SNPs-GRS) may be an easy and practical model available in the clinical practice to assist in identifying the individuals at the highest risk of IgAN because its calculation is simpler compared to the genome-wide polygenic risk score and we could calculate it according to a formula, especially in the absence of well-established risk predictors for predisposing IgAN. Despite the modest discriminative ability, the person in the top 20% of 21 SNPs-GRS have a fourfold increased risk of developing IgAN, suggesting that polygenic risk scores could be utilized in improving screening programs. Similarly, the 21 SNPs-GRS was the most effective in distinguishing high-risk individuals with poor prognosis. The reason may be that the effect sizes of the newly added loci (p < 5E−05) in the 55 SNPs-GRS are mostly less or moderate compared to the 21 SNPs-GRS. Anyhow, more widespread replication in other independent cohorts is needed to verify this result. Since the current GWASs of IgAN only recognize a very limited number of significant variants, the effect of current GRS is limit. Therefore, it is necessary to perform multi-ethnic GWAS with larger sample sizes to identify more significant variants associated with IgA nephropathy. These three GRS models performed poorly in distinguishing high-risk individuals prone to ESKD, while the three GRS models could distinguish individuals at high risk for ESKD or ≥ 50% reduction in eGFR, which is possibly due to the small number of ESKD outcome events. The GRS models based on IgAN susceptibility loci were less effective in distinguishing high-risk individuals prone to poor prognosis than high-risk individuals susceptible to IgAN, which showed that genetic variants and scores linked to susceptibility were likely not all significantly associated with IgAN progression (Liu et al. 2021; Robinson-Cohen et al. 2023). Our results in analyzing the association between single SNPs and disease progression also supported this, with only five disease-susceptible SNPs out of 107 SNPs associated with ESKD as well as nine SNPs associated with ESKD or ≥ 50% reduction in eGFR, and none of the associations survived the multiple-testing correction. Hence, it is necessary to conduct a genome-wide association study related to the adverse prognosis of IgAN to discover genetic loci associated with disease progression, further reveal the mechanism of disease occurrence and progression, and develop targeted drugs to delay disease progression.

For the association between genetic risk scores and clinical characteristics, we found that all three polygenic risk scores were, to various degrees, associated with the early age of biopsy. This suggests the disease may be diagnosed at an earlier time point in those with higher polygenic risk scores. The serum component C3 and IgA/C3 ratio were both associated with 21 SNPs-GRS after Bonferroni correction. Several studies have shown that serum C3 and IgA/C3 ratio were associated with the occurrence and progression of IgAN and maybe the potential prognostic markers (Mizerska-Wasiak 2023), which provide clues for the researchers to study the pathogenic mechanism and also indicated the potential clinical value of genetic risk score to some extent. In addition, we found that both the microhematuria at biopsy and time-averaged microhematuria were associated with the three GRS models. Kiryluk et al. (2023) also found a significant positive correlation between the genome-wide polygenic risk score with hematuria in 590,515 participants using a meta-phenome-wide association study.

The current study includes a large sample size of participants in a regularly followed-up prospective cohort of IgA nephropathy, with strength with the availability of complete genetic information and in-depth clinical data, the comparatively long follow-up period with a certain number of ESKD outcomes available for prospective analysis. We systematically and completely analyzed the relationship between different GRS models and disease susceptibility, clinical characteristics, and prognosis. However, some limitations should also be noted. First, the weights for each SNP used to construct the GRS were derived from the previous GWASs of the IgAN conducted mainly by the cohort of European and Asian populations, which may degrade the performance of the GRS models in the Chinese population due to the non-ethnic-specific effect values of the SNPs. Second, some genetic variations were not used to build GRSs because of the batch effects of using different genotyping arrays and strict quality control, which may also affect the evaluation and comparison of the performance of different GRS models. Finally, its single-center experience, and the inability to generalize to other ancestry groups.

Conclusion

In conclusion, a higher burden of risk variants is associated with earlier disease onset and a higher risk of a worse prognosis. The incorporation of polygenic risk into clinical care settings may provide valuable risk stratification guidance to identify individuals who require regular urinalysis to facilitate early detection of IgAN patients and patient stratification for early intervention to delay disease progression.

Supplementary Information

Below is the link to the electronic supplementary material.

Supplementary file1 (PDF 407 KB) (407.2KB, pdf)
Supplementary file2 (XLSX 89 KB) (88.6KB, xlsx)

Acknowledgements

We thank the National Science Foundation of China, Beijing Natural Science Foundation, Academy of Medical Sciences--Newton Advanced Fellowship, Fok Ying Tung Education Foundation, Chinese Academy of Medical Sciences (CAMS) Innovation Fund for Medical Sciences, and National High-Level Hospital Clinical Research Funding for providing the funding. We also thank all the participants.

Abbreviations

GWASs

Genome-wide association studies

IgAN

IgA nephropathy

GRS

Genetic risk score

SNPs

Single nucleotide polymorphisms

AUC

Area under the receiver operating characteristic curve

BMI

Body mass index

MAP

Mean arterial pressure

ESKD

End-stage kidney disease

GPS

Polygenic risk score

GCKD

German chronic kidney disease

eGFR

Estimated glomerular filtration rate

LD

Linkage disequilibrium

OR

Odds ratio

uwGRS

Unweighted genetic risk score

wGRS

Weighted genetic risk score

SD

Standard deviation

TNFSF13

TNF superfamily member 13

HORMAD2

HORMA domain containing 2

DEFA

Defensin alpha

LYN

LYN proto-oncogene, scr family tyrosine kinase

ODF1

Outer dense fiber of sperm tails 1

KLF10

KLF transcription factor 10

KLHL1

Kelch like family member 1

PTGER4

Prostaglandin E receptor 4

CARD9

Caspase recruitment domain family member 9

MPDU1

Mannose-P-dolichol utilization defect 1

MBP

Myelin basic protein

MTMR3

Myotubularin-related protein 3

Authors’ Contributions

Research idea and study design: XZ, HZ, LX; data acquisition: XZ, HZ, LX, TG, PC, SS; data analysis/interpretation: LX, XZ, HZ; statistical analysis: LX; supervision or mentorship: XZ, HZ, JL, LL, SQ, YL, SS. Each author contributed important intellectual content during manuscript drafting or revision and agrees to be personally accountable for the individual’s own contributions and to ensure that questions pertaining to the accuracy or integrity of any portion of the work, even one in which the author was not directly involved, are appropriately investigated and resolved, including with documentation in the literature if appropriate.

Funding

This work was supported by National Science Foundation of China (82022010, 82370709, 81970613, 82070733, 82000680); Beijing Natural Science Foundation (Z190023); Academy of Medical Sciences–Newton Advanced Fellowship (NAFR13\1033); Fok Ying Tung Education Foundation (171030); Chinese Academy of Medical Sciences (CAMS) Innovation Fund for Medical Sciences (2019-I2M-5–046, 2020-JKCS-009); National High Level Hospital Clinical Research Funding (Interdisciplinary Clinical Research Project of Peking University First Hospital, 2022CR41). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Data availability

The datasets generated during and/or analyzed during the current study are not publicly available due to patient privacy but are available from the corresponding author on reasonable request.

Declarations

Conflict of interest

On behalf of all authors, the corresponding author states that there is no conflict of interest.

Ethical approval

This study was approved by the ethics committee of Peking University First Hospital [IRB number 2020Y197] and was conducted in accordance with the principle of the Helsinki Declaration. Written informed consent was provided by all participants.

Consent to participate

Written informed consent was provided by all participants.

Consent for publication

The participants consent to have their data published.

References

  1. Abraham G, Havulinna AS, Bhalala OG, et al. Genomic prediction of coronary heart disease. Eur Heart J. 2016;37(43):3267–3278. doi: 10.1093/eurheartj/ehw450. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Barsoum RS. Glomerulonephritis in disadvantaged populations. Clin Nephrol. 2010;74(Suppl 1):S44–50. doi: 10.5414/cnp74s044. [DOI] [PubMed] [Google Scholar]
  3. Cattran DC, Coppo R, Cook HT, et al. The Oxford classification of IgA nephropathy: rationale, clinicopathological correlations, and classification. Kidney Int. 2009;76(5):534–545. doi: 10.1038/ki.2009.243. [DOI] [PubMed] [Google Scholar]
  4. Choi SW, O'Reilly PF. PRSice-2: Polygenic Risk Score software for biobank-scale data. GigaScience. 2019 doi: 10.1093/gigascience/giz082. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Dai J, Lv J, Zhu M, et al. Identification of risk loci and a polygenic risk score for lung cancer: a large-scale prospective cohort study in Chinese populations. Lancet Respir Med. 2019;7(10):881–891. doi: 10.1016/s2213-2600(19)30144-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. D'Amico G. The commonest glomerulonephritis in the world: IgA nephropathy. Q J Med. 1987;64(245):709–727. [PubMed] [Google Scholar]
  7. Elliott J, Bodinier B, Bond TA, et al. Predictive accuracy of a polygenic risk score-enhanced prediction model vs a clinical risk score for coronary artery disease. JAMA. 2020;323(7):636–645. doi: 10.1001/jama.2019.22241. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Feehally J, Farrall M, Boland A, et al. HLA has strongest association with IgA nephropathy in genome-wide analysis. J Am Soc Nephrol. 2010;21(10):1791–1797. doi: 10.1681/asn.2010010076. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Floege J, Amann K. Primary glomerulonephritides. Lancet (london, England) 2016;387(10032):2036–2048. doi: 10.1016/s0140-6736(16)00272-5. [DOI] [PubMed] [Google Scholar]
  10. Gharavi AG, Kiryluk K, Choi M, et al. Genome-wide association study identifies susceptibility loci for IgA nephropathy. Nat Genet. 2011;43(4):321–327. doi: 10.1038/ng.787. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Hastings MC, Moldoveanu Z, Julian BA, et al. Galactose-deficient IgA1 in African Americans with IgA nephropathy: serum levels and heritability. Clin J Am Soc Nephrol CJASN. 2010;5(11):2069–2074. doi: 10.2215/cjn.03270410. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Hurson AN, Pal Choudhury P, Gao C, et al. Prospective evaluation of a breast-cancer risk model integrating classical risk factors and polygenic risk in 15 cohorts from six countries. Int J Epidemiol. 2022;50(6):1897–1911. doi: 10.1093/ije/dyab036. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Inouye M, Abraham G, Nelson CP, et al. Genomic risk prediction of coronary artery disease in 480,000 adults: implications for primary prevention. J Am Coll Cardiol. 2018;72(16):1883–1893. doi: 10.1016/j.jacc.2018.07.079. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Julian BA, Quiggins PA, Thompson JS, et al. Familial IgA nephropathy. Evidence of an inherited mechanism of disease. N Engl J Med. 1985;312(4):202–208. doi: 10.1056/nejm198501243120403. [DOI] [PubMed] [Google Scholar]
  15. Kiryluk K, Moldoveanu Z, Sanders JT, et al. Aberrant glycosylation of IgA1 is inherited in both pediatric IgA nephropathy and Henoch-Schönlein purpura nephritis. Kidney Int. 2011;80(1):79–87. doi: 10.1038/ki.2011.16. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Kiryluk K, Li Y, Sanna-Cherchi S, et al. Geographic differences in genetic susceptibility to IgA nephropathy: GWAS replication study and geospatial risk analysis. PLoS Genet. 2012;8(6):e1002765. doi: 10.1371/journal.pgen.1002765. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Kiryluk K, Li Y, Scolari F, et al. Discovery of new risk loci for IgA nephropathy implicates genes involved in immunity against intestinal pathogens. Nat Genet. 2014;46(11):1187–1196. doi: 10.1038/ng.3118. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Kiryluk K, Sanchez-Rodriguez E, Zhou XJ, et al. Genome-wide association analyses define pathogenic signaling pathways and prioritize drug targets for IgA nephropathy. Nat Genet. 2023;55(7):1091–1105. doi: 10.1038/s41588-023-01422-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Kullo IJ, Jouni H, Austin EE, et al. Incorporating a genetic risk score into coronary heart disease risk estimates: effect on low-density lipoprotein cholesterol levels (the MI-GENES clinical trial) Circulation. 2016;133(12):1181–1188. doi: 10.1161/circulationaha.115.020109. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Lai KN, Tang SC, Schena FP, et al. IgA nephropathy. Nat Rev Dis Primers. 2016;2:16001. doi: 10.1038/nrdp.2016.1. [DOI] [PubMed] [Google Scholar]
  21. Lakeman IMM, Hilbers FS, Rodríguez-Girondo M, et al. Addition of a 161-SNP polygenic risk score to family history-based risk prediction: impact on clinical management in non-BRCA1/2 breast cancer families. J Med Genet. 2019;56(9):581–589. doi: 10.1136/jmedgenet-2019-106072. [DOI] [PubMed] [Google Scholar]
  22. Lambert SA, Abraham G, Inouye M. Towards clinical utility of polygenic risk scores. Hum Mol Genet. 2019;28(R2):R133–r142. doi: 10.1093/hmg/ddz187. [DOI] [PubMed] [Google Scholar]
  23. Li M, Yu XQ. Genetic determinants of IgA nephropathy: eastern perspective. Semin Nephrol. 2018;38(5):455–460. doi: 10.1016/j.semnephrol.2018.05.015. [DOI] [PubMed] [Google Scholar]
  24. Li X, Liu Y, Lv J, et al. Progression of IgA nephropathy under current therapy regimen in a Chinese population. Clin J Am Soc Nephrol CJASN. 2014;9(3):484–489. doi: 10.2215/cjn.01990213. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Li M, Foo JN, Wang JQ, et al. Identification of new susceptibility loci for IgA nephropathy in Han Chinese. Nat Commun. 2015;6:7270. doi: 10.1038/ncomms8270. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Li M, Wang L, Shi DC, et al. Genome-wide meta-analysis identifies three novel susceptibility loci and reveals ethnic heterogeneity of genetic susceptibility for IgA nephropathy. J Am Soc Nephrol. 2020;31(12):2949–2963. doi: 10.1681/asn.2019080799. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Liu G, Peng J, Liao Z, et al. Genome-wide survival study identifies a novel synaptic locus and polygenic score for cognitive progression in Parkinson's disease. Nat Genet. 2021;53(6):787–793. doi: 10.1038/s41588-021-00847-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Lu X, Liu Z, Cui Q, et al. A polygenic risk score improves risk stratification of coronary artery disease: a large-scale prospective Chinese cohort study. Eur Heart J. 2022;43(18):1702–1711. doi: 10.1093/eurheartj/ehac093. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Mavaddat N, Michailidou K, Dennis J, et al. Polygenic risk scores for prediction of breast cancer and breast cancer subtypes. Am J Hum Genet. 2019;104(1):21–34. doi: 10.1016/j.ajhg.2018.11.002. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Mizerska-Wasiak M. How to take advantage of easily available biomarkers in patients with IgA nephropathy: IgA and C3 in serum and kidney biopsies. Pediatr Nephrol (berlin, Germany) 2023;38(5):1439–1448. doi: 10.1007/s00467-022-05644-9. [DOI] [PubMed] [Google Scholar]
  31. Natarajan P, Young R, Stitziel NO, et al. Polygenic risk score identifies subgroup with higher burden of atherosclerosis and greater relative benefit from statin therapy in the primary prevention setting. Circulation. 2017;135(22):2091–2101. doi: 10.1161/circulationaha.116.024436. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Neugut YD, Kiryluk K. Genetic determinants of IgA nephropathy: western perspective. Semin Nephrol. 2018;38(5):443–454. doi: 10.1016/j.semnephrol.2018.05.014. [DOI] [PubMed] [Google Scholar]
  33. Northcutt MJ, Shi Z, Zijlstra M, et al. Polygenic risk score is a predictor of adenomatous polyps at screening colonoscopy. BMC Gastroenterol. 2021;21(1):65. doi: 10.1186/s12876-021-01645-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Purcell S, Neale B, Todd-Brown K, et al. PLINK: a tool set for whole-genome association and population-based linkage analyses. Am J Hum Genet. 2007;81(3):559–575. doi: 10.1086/519795. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Roberts IS, Cook HT, Troyanov S, et al. The Oxford classification of IgA nephropathy: pathology definitions, correlations, and reproducibility. Kidney Int. 2009;76(5):546–556. doi: 10.1038/ki.2009.168. [DOI] [PubMed] [Google Scholar]
  36. Robinson-Cohen C, Triozzi JL, Rowan B, et al. Genome-wide association study of CKD progression. J Am Soc Nephrol. 2023 doi: 10.1681/asn.0000000000000170. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Robson ME, Reiner AS, Brooks JD, et al. Association of common genetic variants with contralateral breast cancer risk in the WECARE study. J Natl Cancer Inst. 2017 doi: 10.1093/jnci/djx051. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Schena FP. A retrospective analysis of the natural history of primary IgA nephropathy worldwide. Am J Med. 1990;89(2):209–215. doi: 10.1016/0002-9343(90)90300-3. [DOI] [PubMed] [Google Scholar]
  39. Scolari F, Amoroso A, Savoldi S, et al. Familial clustering of IgA nephropathy: further evidence in an Italian population. Am J Kidney Dis. 1999;33(5):857–865. doi: 10.1016/s0272-6386(99)70417-8. [DOI] [PubMed] [Google Scholar]
  40. Torkamani A, Wineinger NE, Topol EJ. The personal and clinical utility of polygenic risk scores. Nat Rev Genet. 2018;19(9):581–590. doi: 10.1038/s41576-018-0018-x. [DOI] [PubMed] [Google Scholar]
  41. Toulopoulou T, Zhang X, Cherny S, et al. Polygenic risk score increases schizophrenia liability through cognition-relevant pathways. Brain J Neurol. 2019;142(2):471–485. doi: 10.1093/brain/awy279. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Trimarchi H, Barratt J, Cattran DC, et al. Oxford classification of IgA nephropathy 2016: an update from the IgA nephropathy classification working group. Kidney Int. 2017;91(5):1014–1021. doi: 10.1016/j.kint.2017.02.003. [DOI] [PubMed] [Google Scholar]
  43. Tsukamoto Y, Wang H, Becker G, et al. Report of the Asian Forum of Chronic Kidney Disease Initiative (AFCKDI) 2007. “Current status and perspective of CKD in Asia”: diversity and specificity among Asian countries. Clin Exp Nephrol. 2009;13(3):249–256. doi: 10.1007/s10157-009-0156-8. [DOI] [PubMed] [Google Scholar]
  44. Wakai K, Nakai S, Matsuo S, et al. Risk factors for IgA nephropathy: a case-control study with incident cases in Japan. Nephron. 2002;90(1):16–23. doi: 10.1159/000046309. [DOI] [PubMed] [Google Scholar]
  45. Wang YN, Gan T, Qu S, et al. MTMR3 risk alleles enhance toll like receptor 9-induced IgA immunity in IgA nephropathy. Kidney Int. 2023 doi: 10.1016/j.kint.2023.06.018. [DOI] [PubMed] [Google Scholar]
  46. Willer CJ, Li Y, Abecasis GR. METAL: fast and efficient meta-analysis of genomewide association scans. Bioinformatics (oxford, England) 2010;26(17):2190–2191. doi: 10.1093/bioinformatics/btq340. [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Wyatt RJ, Julian BA. IgA nephropathy. N Engl J Med. 2013;368(25):2402–2414. doi: 10.1056/NEJMra1206793. [DOI] [PubMed] [Google Scholar]
  48. Wyatt RJ, Rivas ML, Julian BA, et al. Regionalization in hereditary IgA nephropathy. Am J Hum Genet. 1987;41(1):36–50. [PMC free article] [PubMed] [Google Scholar]
  49. Yang C, Jie W, Yanlong Y, et al. Genome-wide association study identifies TNFSF13 as a susceptibility gene for IgA in a South Chinese population in smokers. Immunogenetics. 2012;64(10):747–753. doi: 10.1007/s00251-012-0636-y. [DOI] [PubMed] [Google Scholar]
  50. Yu XQ, Li M, Zhang H, et al. A genome-wide association study in Han Chinese identifies multiple susceptibility loci for IgA nephropathy. Nat Genet. 2011;44(2):178–182. doi: 10.1038/ng.1047. [DOI] [PubMed] [Google Scholar]
  51. Zhang X, Rice M, Tworoger SS, et al. Addition of a polygenic risk score, mammographic density, and endogenous hormones to existing breast cancer risk prediction models: a nested case-control study. PLoS Med. 2018;15(9):e1002644. doi: 10.1371/journal.pmed.1002644. [DOI] [PMC free article] [PubMed] [Google Scholar]
  52. Zhou XJ, Qi YY, Hou P, et al. Cumulative effects of variants identified by genome-wide association studies in IgA nephropathy. Sci Rep. 2014;4:4904. doi: 10.1038/srep04904. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary file1 (PDF 407 KB) (407.2KB, pdf)
Supplementary file2 (XLSX 89 KB) (88.6KB, xlsx)

Data Availability Statement

The datasets generated during and/or analyzed during the current study are not publicly available due to patient privacy but are available from the corresponding author on reasonable request.

Articles from Phenomics are provided here courtesy of Springer

RESOURCES