Clinical usefulness of the Oxford classification in determining immunosuppressive treatment in IgA nephropathy

Figures & data

Figure 1. Flow diagram of the study. Abbreviation: eGFR: estimated glomerular filtration rate; UPCR: urine protein-to-creatinine ratio.

Table 1. Baseline characteristics according to the Oxford classification (M and E, n = 377).

Variables	M0 (n = 303, 80.4%)	M1 (n = 74, 19.6%)	p	E0 (n = 292, 77.5%)	E1 (n = 85, 22.5%)	p
Age (years)	38.7 ± 13.0	37.6 ± 12.9	.52	38.3 ± 12.7	38.8 ± 14.0	.76
Sex (male, %)	143 (57.2)	37 (50.0)	.67	141 (48.3)	39 (45.9)	.70
MAP (mmHg)	91.2 ± 12.0	91.0 ± 13.1	.88	91.2 ± 12.2	90.9 ± 12.5	.83
Body mass index (kg/m^2)	23.0 ± 3.6	22.5 ± 3.7	.42	22.8 ± 3.5	23.4 ± 4.0	.23
Hypertension (%)	45 (17.4)	29 (24.4)	.12	54 (20.9)	31 (26.1)	.27
BUN (mg/dl)	13.7 ± 4.4	14.7 ± 5.0	.15	13.8 ± 4.5	14.4 ± 4.5	.25
Creatinine (mg/dl)	0.86 ± 0.24	0.96 ± 0.27	.01	0.87 ± 0.23	0.92 ± 0.29	.09
eGFR (ml/min/1.73 m^2)	109.0 ± 41.1	97.6 ± 35.7	.03	108.7 ± 41.9	100.2 ± 33.6	.09
UPCR (g/g Cr)	0.54 (0.31–0.75)	0.67 (0.38–0.82)	.01^a	0.53 (0.30–0.74)	0.65 (0.48–0.82)	.01^a
RASB (%)	224 (73.9)	68 (91.9)	.01	211 (72.3)	77 (90.6)	.01
IgA (mg/dl)	323.3 ± 114.3	338.9 ± 111.5	.48	328.4 ± 115.1	319.0 ± 111.0	.59
Hemoglobin (g/dl)	13.2 ± 1.6	13.1 ± 1.6	.78	13.2 ± 1.6	13.1 ± 1.5	.66
Glucose (mg/dl)	95.1 ± 12.7	93.6 ± 10.7	.48	95.2 ± 13.0	93.6 ± 10.3	.35
Serum albumin (g/dl)	4.1 ± 0.4	4.0 ± 0.4	.06	4.1 ± 0.4	3.9 ± 0.4	.01
Cholesterol (mg/dl)	184.5 ± 37.9	188.9 ± 36.5	.39	182.1 ± 35.6	196.2 ± 42.1	.01
Triglyceride (mg/dl)	96 (69–148)	104 (79–163)	.23^a	94 (70–147)	108 (74–167)	.25^a
HDL-C (mg/dl)	52.4 ± 14.0	57.5 ± 17.6	.13	52.3 ± 13.8	55.1 ± 16.2	.25
LDL-C (mg/dl)	114.0 ± 32.6	111.7 ± 30.2	.73	109.4 ± 30.5	123.9 ± 34.0	.01
hs-CRP (mg/L)	1.27 (0.34–4.50)	1.30 (0.45–4.59)	.77^a	1.12 (0.10–2.87)	2.12 (0.93–5.58)	.03^a

Note: Variables are expressed as median (range), mean ± SD, or n (%); ^aMann–Whitney U test.

Abbreviations: MAP: mean arterial pressure; BUN: blood urea nitrogen; eGFR: estimated glomerular filtration rate; UPCR: urine protein-to-creatinine ratio; RASB: renin-angiotensin system blockade; HDL-C: high-density lipoprotein cholesterol; LDL-C: low-density lipoprotein cholesterol; hs-CRP: high sensitivity C-reactive protein.

Table 2. Baseline characteristics according to the Oxford classification (S and T, n = 377).

Variables	S0 (n = 146, 38.7%)	S1 (n = 231, 61.3%)	p	T0 (n = 337, 89.4%)	T1 (n = 40, 10.6%)	p
Age (years)	38.0 ± 13.8	38.7 ± 12.5	.61	37.8 ± 12.9	44.1 ± 12.5	.01
Sex (male, %)	75 (51.4)	105 (45.5)	.26	155 (46.0)	25 (62.5)	.05
MAP (mmHg)	90.4 ± 12.2	91.7 ± 12.2	.32	90.6 ± 12.0	96.2 ± 13.3	.01
Body mass index (kg/m^2)	22.5 ± 3.7	23.1 ± 3.6	.21	22.8 ± 3.7	24.1 ± 3.1	.18
Hypertension (%)	145 (56.2)	86 (72.3)	.01	25 (9.7)	15 (12.6)	.39
BUN (mg/dl)	13.5 ± 4.9	14.2 ± 4.3	.19	13.5 ± 4.0	17.3 ± 6.8	.01
Creatinine (mg/dl)	0.86 ± 0.24	0.89 ± 0.25	.22	0.86 ± 0.23	1.06 ± 0.29	<.001
eGFR (ml/min/1.73 m^2)	112.0 ± 46.2	103.4 ± 35.9	.04	109.1 ± 40.1	87.2 ± 37.6	.01
UPCR (g/g Cr)	0.46 (0.18–0.70)	0.62 (0.39–0.81)	<.001^a	0.55 (0.31–0.76)	0.68 (0.41–0.86)	.01^a
RASB total (%)	94 (64.4)	191 (82.7)	.01	254 (75.4)	40 (100.0)	.03^b
IgA (mg/dl)	300.8 ± 116.2	336.7 ± 111.2	.04	322.3 ± 113.5	397.8 ± 97.4	.05
Hemoglobin (g/dl)	13.3 ± 1.6	13.1 ± 1.6	.56	13.2 ± 1.6	13.3 ± 1.8	.70
Glucose (mg/dl)	97.3 ± 16.7	93.7 ± 9.7	.03	94.4 ± 11.9	101.7 ± 18.4	.15
Serum albumin (g/dl)	4.1 ± 0.5	4.0 ± 0.4	.01	4.0 ± 0.4	4.0 ± 0.4	.26
Cholesterol (mg/dl)	177.6 ± 35.4	190.2 ± 38.3	.01	184.0 ± 36.9	197.5 ± 42.4	.04
Triglyceride (mg/dl)	98 (65–139)	97 (74–154)	.16^a	95 (70–147)	136 (75–225)	.07^a
HDL-C (mg/dl)	50.5 ± 12.9	54.2 ± 15.2	.12	52.9 ± 14.4	55.6 ± 16.8	.52
LDL-C (mg/dl)	106.0 ± 28.9	117.1 ± 33.0	.03	113.8 ± 32.2	112.1 ± 32.6	.86
hs-CRP (mg/L)	1.20 (0.10–4.24)	1.34 (0.45–4.55)	.71^a	1.24 (0.38–4.31)	2.08 (0.75–16.30)	.32^a

Note: Variables are expressed as median (range), mean ± SD, or n (%); ^aMann–Whitney U test, ^bFisher’s exact test.

Abbreviations: MAP: mean arterial pressure; BUN: blood urea nitrogen; eGFR: estimated glomerular filtration rate; UPCR: urine protein-to-creatinine ratio; RASB: renin-angiotensin system blockade; HDL-C: high-density lipoprotein cholesterol; LDL-C: low-density lipoprotein cholesterol; hs-CRP: high sensitivity C-reactive protein.

Figure 2. Kaplan–Meier plots for the development of persistent proteinuria according to the Oxford-MEST.

Figure 3. Time course of UPCR according to Oxford classification. Each point was demonstrated as the mean values with standard errors of UPCR on every 6-month follow-up. Abbreviation: UPCR: urine protein-to-creatinine ratio.

Table 3. Cox regression model with time-dependent covariates for the association of Oxford classification with development of persistent proteinuria (n = 377).

M1 lesion	HR (95% CI)	p
Crude	3.039 (1.69–5.48)	<.001
Adjusted
Model 1^a	2.643 (2.22–3.15)	<.001
Model 2^b	2.381 (1.98–2.87)	<.001

^a Model 1: adjusted for age, sex, MAP, UPCR, and eGFR.

^b Model 2: adjusted for age, sex, MAP, UPCR, eGFR, E, S, and T.

Abbreviations: HR: hazard ratio; CI: confidence interval; MAP: mean arterial pressure; UPCR: urine protein-to-creatinine ratio; eGFR: estimated glomerular filtration rate.

Figure 4. Receiver operating characteristics curve for the development of persistent proteinuria by the Oxford-MEST.

Table 4. C-Statistics for prediction of the development of persistent proteinuria using multivariate Cox’s regression models.

		p for difference of C-statistics compared with models
Model	C-statistics (95% CI)	Model 1^a	Model 2^b	Model 3^c
Model 1^a	0.703 (0.616–0.791)	NA	–	–
Model 2^b	0.666 (0.574–0.758)	0.02	NA	–
Model 3^c	0.663 (0.571–0.754)	0.01	0.52	NA
Model 4^d	0.665 (0.574–0.756)	0.03	0.61	0.11

^a Model 1: Age, sex, MAP, UPCR, eGFR, RASB, glucocorticoid treatment, and M.

^b Model 2: Age, sex, MAP, UPCR, eGFR, RASB, glucocorticoid treatment, and E.

^c Model 3: Age, sex, MAP, UPCR, eGFR, RASB, glucocorticoid treatment, and S.

^d Model 4: Age, sex, MAP, UPCR, eGFR, RASB, glucocorticoid treatment, and T.

Abbreviations: CI: confidence interval; NA: not applicable; MAP: mean arterial pressure; UPCR: urine protein-to-creatinine ratio; eGFR: estimated glomerular filtration rate; RASB: renin-angiotensin system blockade.

Table 5. Cox proportional hazard regression analysis for the association of Oxford classification with a 30% decline in eGFR (n = 377).

M1	Hazard ratio (95% CI)	p
Crude	1.250 (0.676–2.314)	.48
Adjusted^a	3.546 (1.189–10.576)	.02

^a Model was adjusted for age, sex, MAP, eGFR, UPCR, RASB, and glucocorticoid treatment.

Abbreviations: CI: confidence interval; eGFR: estimated glomerular filtration rate; MAP: mean arterial pressure; UPCR: urine protein-to-creatinine ratio; RASB: renin-angiotensin system blockade.

Figure 5. The changes in UPCR after 6 months with (A) and without (B) glucocorticoids treatment. Abbreviation: UPCR: urine protein-to-creatinine ratio.

Figure 6. A Kaplan–Meier plot and eGFR decline rate in the matched cohort of glucocorticoid users versus nonusers with M1 (A). The mean with standard errors were used in time course of eGFR (B) and UPCR figures (C). Abbreviation: eGFR: estimated glomerular filtration rate; UPCR: urine protein-to-creatinine ratio.

Table 6. Baseline characteristics in the unmatched and matched cohort according to glucocorticoids treatment.

	Unmatched M1 (n = 58)			Matched M1 (n = 36)
Variables	Nonuser (n = 37, 63.8%)	User (n = 21, 36.2%)	p	Nonuser (n = 18, 50.0%)	User (n = 18, 50.0%)	p
Age (years)	40.5 ± 12.3	41.3 ± 13.0	.84	41.3 ± 12.0	41.4 ± 13.9	.9
Sex (male, %)	17 (46.0)	11 (52.4)	.43	8 (44.4)	9 (50.0)	.74
MAP (mmHg)	92.4 ± 13.8	93.4 ± 14.1	.82	96.1 ± 14.5	91.7 ± 14.6	.37
Creatinine (mg/dl)	0.91 ± 0.26	1.16 ± 0.26	.01	1.11 ± 0.30	1.13 ± 0.28	.80
eGFR (ml/min/1.73 m^2)	98.8 ± 37.2	72.7 ± 22.1	.01	76.7 ± 25.6	75.5 ± 26.2	.89
UPCR (g/g Cr)	1.69 (1.33–2.09)	2.29 (1.34–2.74)	.11^a	1.42 (0.58–2.08)	1.65 (0.71–2.64)	.28^a
IgA (mg/dl)	355.9 ± 87.0	317.2 ± 84.6	.34	343.1 ± 88.9	325.2 ± 89.5	.51
Hemoglobin (g/dl)	13.2 ± 1.8	12.6 ± 1.4	.38	12.8 ± 1.9	12.7 ± 1.6	.39
Serum albumin (g/dl)	3.9 ± 0.3	3.7 ± 0.5	.16	3.8 ± 0.4	3.7 ± 0.5	.45
Cholesterol (mg/dl)	203.7 ± 40.0	204.3 ± 31.8	.9	201.5 ± 44.7	205.1 ± 34.1	.9
hs-CRP (mg/L)	1.34 (0.47–16.37)	2.89 (1.15–48.50)	.52^a	2.34 (0.98–24.65)	2.84 (1.66–39.10)	.73^a

Note: Variables are expressed as median (range), mean ± SD, or n (%); ^aMann–Whitney U test.

Abbreviations: MAP: mean arterial pressure; eGFR: estimated glomerular filtration rate; UPCR: urine protein-to-creatinine ratio; hs-CRP: high sensitivity C-reactive protein.

Table 7. Cox proportional hazard analysis for the effect of glucocorticoid treatment on a 30% decline in eGFR with the unmatched and matched M1 cohort.

	Unmatched M1 (n = 58)		Matched M1 (n = 36)
	HR (95% CI)	p	HR (95% CI)	p
Crude	2.604 (0.82–8.24)	.10	2.006 (0.58–6.90)	.27
Model 1^a	3.081 (0.94–8.10)	.06	1.992 (0.57–6.96)	.28
Model 2^b	1.177 (0.35–4.01)	.79	2.216 (0.64–7.73)	.21
Model 3^c	2.551 (0.78–8.37)	.12	1.932 (0.55–6.78)	.30

^a Model 1: adjusted for glucocorticoid treatment, age, and sex.

^b Model 2: adjusted for glucocorticoid treatment, RASB, and eGFR.

^c Model 3: adjusted for glucocorticoid treatment, UPCR, and MAP.

Abbreviations: eGFR: estimated glomerular filtration rate; HR: hazard ratio; CI: confidence interval; RASB: renin-angiotensin system blockade; UPCR: urine protein-to-creatinine ratio; MAP: mean arterial pressure.