Clinical significance of glomerular IgM deposit in IgA nephropathy: a 5-year follow-up study

Figures & data

Figure 1. Flowchart of case ascertainment.

Table 1. baseline characteristics of patients according to the degree of IgM deposits in renal biopsy after matching.

Variables	IgM-H	IgM-L	p
N	266	266
Male	96 (36)	107 (40)	0.43
Age, y	41 (29, 49)	39 (32, 49)	0.82
Systolic blood pressure, mmHg	116.49 (108.29, 127.81)	121.33 (112.58, 132.22)	< 0.01
Diastolic blood pressure, mmHg	73.1 (67.17, 77.62)	76 (69.01, 83)	< 0.01
Body mass index	22.68 (20.67, 25.24)	23.46 (20.81, 25.81)	0.31
Hemoglobin, g/L	125 (114, 138)	126 (115.12, 141)	0.33
Serum albumin, g/L	35.7 (33.3, 38.7)	36.7 (33.9, 39.85)	0.04
Serum creatinine, mmol/L	84.5 (66, 110)	81 (64, 109.75)	0.39
eGFR	83.84 (63.83, 106.81)	90.09 (63.43, 111.81)	0.31
Uric acid, mmol/L	360.32 ± 92.26	362.62 ± 93.2	0.77
Total cholesterol, mmol/L	4.92 (4.24, 5.82)	5 (4.33, 5.78)	0.72
Triglyceride, mmol/L	1.63 (1.18, 2.4)	1.62 (1.08, 2.21)	0.21
HDL-cholesterol, mmol/L	1.06 (0.88, 1.3)	1.11 (0.95, 1.32)	0.06
LDL-cholesterol, mmol/L	2.79 (2.2, 3.32)	3 (2.54, 3.6)	< 0.01
SII	6.21 (4.49, 8.05)	5.18 (3.95, 7.76)	<0.01
Serum IgA, g/L	3.12 (2.52, 3.88)	3.2 (2.55, 3.95)	0.81
Serum IgM, g/L	10.8 (9.18, 12.8)	10.1 (7.94, 12.28)	< 0.01
Serum IgG, g/L	1.25 (0.87, 1.57)	1.06 (0.71, 1.47)	< 0.01
Serum C3, g/L	1.07 (0.94, 1.23)	1.06 (0.89, 1.21)	0.08
Serum C4, g/L	0.25 (0.2, 0.3)	0.23 (0.19, 0.28)	0.03
24h urinary protein, g/d	1.11 (0.56, 2.48)	0.95 (0.43, 1.82)	0.02
UACR, mg/g	491.46 (232.27, 1020.77)	324.94 (114.1, 905.83)	<0.01
Treatment
RAS blocker	225 (85)	229 (86)	0.73
Glucocorticoid	140 (53)	126 (47)	0.29
Immunosuppressant	78 (29)	53 (20)	0.02
Cyclophosphamide	16 (6.0)	17 (6.4)	0.97
Mycophenolate Mofetil	19 (7.1)	17 (6.4)	0.91
Tripterygium Wilfordii	47 (17.7)	26 (9.8)	<0.01
Pathology
M1	84 (44)	128 (60)	<0.01
E1	98 (51)	79 (37)	<0.01
S1	146 (76)	138 (65)	0.02
T-score			0.07
1	51 (26)	42 (20)
2	16 (8)	10 (5)
C1	148 (56)	143 (55)	0.98
IgA			< 0.01
1+	1 (0.4)	1 (0.4)
2+	11 (4)	24 (9)
3+	179 (67)	205 (77)
4+	75 (28)	36 (14)
IgM			< 0.01
1+	0 (0)	266 (100)
>1+	266 (100)	0 (0)
IgG			0.31
1+	11 (4)	16 (6)
2+	23 (9)	13 (5)
3+	3 (1)	3 (1)
C3			0.04
1+	1 (0.4)	1 (0.4)
2+	11 (5)	24 (9)
3+	254 (95)	241 (91)
C4(1+)	5 (2)	0 (0)	0.06
Follow up
Follow-up duration, m	34.97 (19.1, 48.6)	28.43 (16.03, 48.63)	0.22
TA SII	5.38 (3.98, 7.39)	5.13 (3.85, 6.68)	0.18
TA UACR, mg/g	411.31 (183.24, 916.87)	264.97 (106.07, 717.45)	<0.01
Composite endpoint	23 (9)	7 (3)	<0.01

Numerical variables are reported as mean ± SD or median (IQR), while categorical variables are presented as counts (%).

IgM-H refers to IgAN patients whose IF intensity of IgM deposits exceeded 1+, while IgM-L represents IgAN patients whose IF intensity of IgM deposits was equal to 1+. The composite endpoint was referred to a continuous decline in estimated eGFR from baseline of ≥50% or reaching ESRD.

E; endocapillary hypercellularity; M1; mesangial hypercellularity; T1-2: interstitial fibrosis/tubular atrophy; S: segmental glomerulosclerosis/adhesion; C1: crescent; UACR: urine albumin/creatinine ratio; SII: systemic immune-inflammation index; TA UACR: time averaged urine albumin/creatinine ratio; TA SII: time averaged immune-inflammation index.

p-value was adjusted for multiple comparisons.

Figure 2. (A) Disparities in clinical data between the IgM-L and IgM-H in different subgroups. The subgroups were constructed according to the pathologic manifestation in the Oxford classification. The medians, interquartile ranges, and distributions of clinical data in each subgroup were calculated to perform the violin, scatter, and box plots. The p value <0.05 was marked as ‘*’ and p < 0.01 as ‘**’ in the plot. (B) The network diagram illustrates the clustering relationship of clinical factors within IgM-L and IgM-H. All linear data were converted to hierarchical data according to the quantile. Each factor is represented by a node, with circle nodes representing IgM-H and square nodes representing IgM-L. Nodes that are linked by solid black lines indicate a stronger correlation and form a cluster, while nodes that are linked by solid gray lines represent the same factors from various groups (assigned a small weight for the same factors and a larger weight for different factors).

IgM-H refers to IgAN patients whose IF intensity of IgM deposits exceeded 1+, while IgM-L represents IgAN patients whose IF intensity of IgM deposits was equal to 1+. Alb: serum albumin; BMI: body mass index; Hb: hemoglobin; HDL: HDL-cholesterol; LDL: LDL-cholesterol; SBP: systolic blood pressure; DBP: diastolic blood pressure; S-Cr: serum creatinine; TC: total cholesterol; TG: triglyceride; UA: serum uric acid; S-IgA: serum IgA; S-IgG: serum IgG; S-IgM: serum IgM; S-C3: serum C3; S-C4: serum C4; UACR: urine albumin/creatinine ratio; TA UACR: time averaged urine albumin/creatinine ratio; SII: systemic immune-inflammation index; TA SII: time averaged systemic immune-inflammation index; 24h-UP: 24-hour urine protein.

Figure 3. (A) Kaplan-Meier Curve displayed the prognosis of the IgM + and IgM– cohorts (p = 0.27 [log-rank test]). The endpoint was defined as either a reduction of ≥50% in baseline eGFR or the development of ESRD. (B) Estimated marginal means and corresponding SEs of serum creatinine levels were estimated and compared between IgM + and IgM– cohorts within the follow-up period using GLMM. (C) Estimated marginal means and corresponding SEs of urine albumin/creatinine ratio levels were estimated and compared between IgM + and IgM– cohorts within the follow-up period using GLMM. (D) Estimated marginal means and corresponding SEs of serum albumin levels were estimated and compared between IgM + and IgM– cohorts within the follow-up period using GLMM.

Statistical significance levels are indicated in the plot as ‘*’, which represents p < 0.05, and ‘**’, representing p < 0.01. The ANOVA results of the GLMM are presented in Table S6.

IgM + refers to IgAN patients with the presence of IgM deposits. IgM– refers to IgAN patients with the absence of IgM deposits. Alb: serum albumin; S-Cr: serum creatinine; UACR: urine albumin/creatinine ratio.

Figure 4. (A) Kaplan-Meier curve displayed the prognosis of the IgM-L and IgM-H cohorts (p = 0.02 [log-rank test]). The endpoint was defined as either a reduction of ≥50% in baseline eGFR or the development of ESRD. (B) Estimated marginal means and corresponding SEs of serum creatinine levels were estimated and compared between IgM-L and IgM-H cohorts within the follow-up period using GLMM. (C) Estimated marginal means and corresponding SEs of urine albumin/creatinine ratio levels were estimated and compared between IgM-L and IgM-H cohorts within the follow-up period using GLMM. (D) Estimated marginal means and corresponding SEs of serum albumin levels were estimated and compared between IgM-L and IgM-H cohorts within the follow-up period using GLMM.

Statistical significance levels are indicated in the plot as ‘*’, which represents p < 0.05, and ‘**’, representing p < 0.01. The ANOVA results of the GLMM are presented in Table S7.

IgM-H refers to IgAN patients whose IF intensity of IgM deposits exceeded 1+, while IgM-L represents IgAN patients whose IF intensity of IgM deposits was equal to 1+. Alb: serum albumin; S-Cr: serum creatinine; UACR: urine albumin/creatinine ratio.

Table 2. Factors associated with reaching the composite endpoint in matched IgM-L and IgM-H cohort.

Variable	HR	95% CI	p	HR	95% CI	p
Age	0.97	0.93–1	0.03	0.97	0.94–1.01	0.21
Male	1.78	0.86–3.69	0.1
BMI	1.00	1–1.01	0.1
MAP	1.53	1.24–1.88	<0.01	1.54	1.12–2.13	<0.01
Hemoglobin	0.78	0.63–0.95	0.02
Albumin	0.90	0.85–0.96	<0.01	0.99	0.9–1.08	0.82
Serum creatinine	3.89	2.8–5.43	<0.01	3.05	1.87–4.99	<0.01
Uric acid	1.29	1.03–1.62	0.03
Total cholesterol	1.21	0.96–1.51	0.1
Triglyceride	1.18	1–1.38	0.04	0.77	0.54–1.11	0.22
HDL-cholesterol	0.17	0.04–0.8	0.03
LDL-cholesterol	1.39	1–1.95	0.06
TA SII	1.03	0.93–1.14	0.5
TA UACR	1.13	1.1–1.17	<0.01	1.15	1.1–1.2	<0.01
24h-UP	1.39	1.26–1.53	<0.01
M1	4.61	1.29–16.51	0.02
E1	1.44	0.52–3.98	0.5
S1	1.37	0.44–4.27	0.6
T-score
T0	Reference	Reference	–
T1	7.17	1.39–27.08	0.02
T2	9.19	2.8–32.22	<0.01
C1	0.84	0.42–1.74	0.6
IF intensity of IgA	1.47	0.92–4.13	0.9
IF intensity of IgG	0.64	0.32–2.35	0.8
IF intensity of C3	1.44	0.43–1.86	0.9
Serum IgA	0.89	0.62–1.29	0.5
Serum IgM	0.88	0.78–0.99	0.03
Serum IgG	1.51	0.86–2.65	0.2
Serum C3	1.54	0.3–8.02	0.6
Serum C4	0.82	0.5–1.98	0.1
Decreased intensity of IgM deposits	0.39	0.17–0.92	0.03	0.38	0.16–0.90	0.03
Glucocorticoid	1.14	0.54–2.39	0.7
Immunosuppressant	1.18	0.54–2.61	0.7
RAS blocker	0.67	0.27–1.64	0.4

IgM-H refers to IgAN patients whose IF intensity of IgM deposits exceeded 1+, while IgM-L represents IgAN patients whose IF intensity of IgM deposits was equal to 1+. Composite endpoint was defined as eGFR decreasing from the baseline ≥50% continuously or reaching ESRD. Factors with statistical differences (p < 0.1) in univariate Cox regression were selected to build a multivariate Cox regression model. Then, the AIC using the forward-backward stepwise was performed to select the optimal model with the lowest AIC value. BMI: body mass index; TA UACR: time averaged urine albumin/creatinine ratio; TA SII: time averaged systemic immune-inflammation index; MAP: mean arterial pressure; M1: mesangial hypercellularity; E1: endocapillary hypercellularity; S1: segmental glomerulosclerosis/adhesion; T1-2: the severity of tubular atrophy/interstitial fibrosis; C1: presence of crescent; HR: hazard ratio; CI: confidence interval; AIC: Akaike information criterion; 24h-UP: 24-h urine protein.

Table 3. Association between intensity of glomerular IgM deposit and composite endpoint (multivariate Cox regression model).

Variables	Model1		Model2		Model3		Model4
	HR (95%CI)	p	HR (95%CI)	p	HR (95%CI)	p	HR (95%CI)	p
Intensity of IgM deposit
1+	Ref		Ref		Ref		Ref
>1+	2.63 (1.11–6.25)	0.03	2.31 (1.04–6.66)	0.04	2.32 (1.04–6.64)	0.04	2.57 (1.07–5.68)	0.02

Composite endpoint was defined as eGFR decreasing from the baseline ≥50% continuously or reaching ESRD. HR: Hazard Ratio; CI: Confidence Interval; M1: mesangial hypercellularity; E1: endocapillary hypercellularity; S1: segmental glomerulosclerosis/adhesion; T1-2: the severity of tubular atrophy/interstitial fibrosis; C1: presence of crescent.

Model1: Adjust for gender, age, follow-up time and eGFR.

Model2: Model1 + Adjust for M, E, S, T, C.

Model3: Model2 + Adjust for glomerular IgA, IgG, C3 deposits.

Model4: Model3 + Adjust for use of glucocorticoid, immunosuppressant and RAS blocker.

Data availability statement

Data is not publicly available due to ethical reasons. Further enquiries can be directed to the corresponding author.