Decoy Receptor 3, a Novel Inflammatory Marker, and Mortality in Hemodialysis Patients

Szu-Chun Hung; Ta-Wei Hsu; Yao-Ping Lin; Der-Cherng Tarng

doi:10.2215/CJN.08410811

Article Figures & Data

Figures

Tables

Download figure
Open in new tab
Download powerpoint
Figure 1.
DcR3 correlates with inflammatory markers. Univariate analysis of the correlation of DcR3 with (A) IL-6, (B) hs-CRP, (C) serum albumin, (D) ICAM-1, and (E) VCAM-1. Logarithmic transformation of DcR3, IL-6, and hs-CRP was used to normalize the distributions for univariate analyses. DcR3, decoy receptor 3; hs-CRP, high-sensitivity C-reactive protein; ICAM-1, intercellular adhesion molecule-1; VCAM-1, vascular cell adhesion molecule-1.
Download figure
Open in new tab
Download powerpoint
Figure 2.
Predictive accuracy of DcR3, IL-6, and albumin estimated by using time-dependent ROC curve analysis. A shows the time-dependent AUCs for all of the markers and B (DcR3), C (IL-6), and D (albumin) show the ROC curves at different time periods. The AUCs at 36 months for DcR3, IL-6, and albumin were 0.74, 0.66 (P<0.01 versus DcR3), and 0.66 (P<0.01 versus DcR3), respectively. The AUCs at 48 month for DcR3, IL-6, and albumin were 0.73, 0.65 (P<0.01 versus DcR3), and 0.68 (P<0.05 versus DcR3), respectively. AUC, area under curve; DcR3, decoy receptor 3; ROC, receiver operating characteristic.
Download figure
Open in new tab
Download powerpoint
Figure 3.
Kaplan–Meier analysis curves in hemodialysis patients at risk for cardiovascular and all-cause mortality. All patients were stratified by the tertiles of baseline serum decoy receptor 3 (DcR3) to assess the unadjusted risks for (A) cardiovascular mortality and (B) all-cause mortality.

Tables

Figures

View popup

Table 1.

Patients’ characteristics according to tertiles of DcR3 levels

Characteristic	Serum DcR3 Tertiles			P Value
Characteristic	0.05−0.92 ng/ml (n=105)	0.93−2.41 ng/ml (n=105)	2.45−17.78 ng/ml (n=106)	P Value
Age (yr)	58±12	59±14	61±13	0.06^a
Male sex	48 (45.7)	52 (49.5)	50 (47.2)	0.86^b
Smoking history	35 (33.3)	32 (30.5)	35 (33.0)	0.89^b
Hypertension	47 (44.8)	46 (43.8)	52 (49.1)	0.72^b
Diabetes mellitus	33 (31.4)	36 (34.3)	35 (33.0)	0.91^b
Previous CVD	15 (14.3)	21 (20.0)	30 (28.6)	0.04^b
Hemodialysis vintage (mo)	72 (37–131)	76 (36–120)	61 (26–118)	0.34^c
Body mass index (kg/m²)	22.1±3.4	22.3±3.1	22.1±4.2	0.87^a
Systolic BP (mmHg)	134±26	134±22	138±24	0.40^a
Diastolic BP (mmHg)	75±15	75±10	76±11	0.73^a
RAS blockade	27 (25.7)	30 (28.6)	29 (27.3)	0.90^b
Total no. of antihypertensives	3 (0–4)	3 (0–3)	4 (0–5)	0.67^c
Statin	16 (15.2)	15 (14.3)	19 (17.9)	0.75^b
Laboratory parameters
Kt/V urea	2.10±0.48	2.05±0.41	2.01±0.74	0.58^a
albumin (g/dl)	4.01±0.31	3.95±0.41	3.79±0.35	<0.001^a
hs-CRP (mg/L)	3.75 (1.30–6.39)	3.86 (1.52–8.24)	5.70 (2.23–8.81)	0.01^c
IL-6 (pg/ml)	2.79 (1.66–5.53)	4.09 (2.20–4.09)	5.36 (2.52–14.4)	<0.001^c
ICAM-1 (ng/ml)	431 (333–599)	511 (385–738)	668 (437–818)	0.02^c
VCAM-1 (ng/ml)	1903 (1491–2391)	2139 (1585–2798)	2550 (1672–3771)	0.04^c
plasma glucose (mg/dl)	113±28	112±29	110±29	0.90^a
total cholesterol (mg/dl)	193±48	192±45	192±47	0.90^a
triglyceride (mg/dl)	190 (182–248)	188 (181–260)	189 (184–237)	0.88^c
calcium (mg/dl)	9.7±0.7	9.6±0.9	9.6±0.8	0.62^a
phosphate (mg/dl)	5.1±1.3	5.2±1.5	4.9±1.3	0.21^a
intact PTH (pg/ml)	254 (109–421)	217 (94–530)	202 (65–468)	0.81^c
hemoglobin (g/dl)	10.6±1.5	10.6±1.6	10.3±1.6	0.30^a
dose of epoetin (U/kg per week)	71 (30–99)	68 (33–94)	71 (20–95)	0.67^c
ferritin (μg/L)	288 (160–452)	322 (182–492)	346 (196–585)	0.29^c
transferrin saturation (%)	25±12	25±11	26±14	0.68^a

All variables were expressed as n (%) for categorical data and as means ± SD or medians and interquartile ranges for continuous data with or without a normal distribution, respectively. Previous CVD category consisted of coronary artery disease, cerebrovascular disease, and peripheral arterial disease. DcR3, decoy receptor 3; CVD, cardiovascular disease; RAS, renin-angiotensin system; hs-CRP, high-sensitivity C-reactive protein; ICAM-1, intercellular adhesion molecule-1; VCAM-1, vascular cell adhesion molecule-1; PTH, parathyroid hormone.
↵a Statistical analysis by ANOVA.
↵b Statistical analysis by the Pearson chi-squared test.
↵c Statistical analysis by the Kruskal–Wallis test.

View popup

Table 2.

Multivariate regression models for the prediction of DcR3

Parameter	Difference in DcR3	SEM	P Value
Age (yr)	−0.004	0.013	0.75
Sex (male)	0.014	0.058	0.81
Diabetes mellitus (presence)	0.008	0.073	0.42
Previous CVD (presence)	0.165	0.061	0.04
HD vintage (mo)	−0.009	0.018	0.69
Kt/V urea	−0.036	0.021	0.28
log IL-6 (pg/ml)	0.248	0.071	0.26
log ICAM-1 (ng/ml)	0.134	0.072	0.07
log VCAM-1 (ng/ml)	0.202	0.096	0.01
Albumin (g/dl)	−0.229	0.087	0.009
Body mass index (kg/m²)	−0.016	0.009	0.52

The adjusted r² of the model was 0.294. Logarithmic transformation of DcR3, IL-6, hs-CRP, ICAM-1 and VCAM-1 was used to normalize the distributions for multivariate analysis. DcR3, decoy receptor 3; CVD, cardiovascular disease; HD, hemodialysis; hs-CRP, high-sensitivity C-reactive protein; ICAM-1, intercellular adhesion molecule-1; VCAM-1, vascular cell adhesion molecule-1.

View popup

Table 3.

Prediction model of pertinent factors for mortality using AUC

Variable	AUC
Traditional cardiovascular risk factors	0.75 (0.70−0.80)
VCAM-1	0.76 (0.71−0.80)
VCAM-1 + albumin	0.78 (0.74−0.83)
VCAM-1 + albumin + IL-6	0.79 (0.75−0.84)
VCAM-1 + albumin + IL-6 + DcR3	0.80 (0.75−0.84)

Risk prediction was assessed by the C statistic. Each variable was stepwise added to assess the incremental change in AUC for predicting mortality at 48 months. Traditional cardiovascular risk factors included age, sex, smoking status, diabetes, total cholesterol, and systolic BP in the model. AUC, area under the ROC curve; VCAM-1, vascular cell adhesion molecule-1; DcR3, decoy receptor 3.

View popup

Table 4.

Multivariate Cox proportional hazards analysis for relative risk of cardiovascular and overall mortality calculated for DcR3 tertiles and for a 1-SD unit change in log DcR3 levels in a follow-up of 54 months

	Cardiovascular Mortality			All-Cause Mortality
	Hazard Ratio (95% Confidence Interval)			Hazard Ratio (95% Confidence Interval)
	Crude	Age- and Sex-Adjusted	Multivariate Adjustment	Crude	Age- and Sex-Adjusted	Multivariate Adjustment
DcR3 by tertiles
lower tertile	1.0	1.0	1.0	1.0	1.0	1.0
middle tertile	1.4 (0.5−3.9)	1.3 (0.4−3.6)	1.3 (0.5−3.8)	1.6 (0.9−2.8)	1.5 (0.8−2.6)	1.4 (0.9−2.5)
upper tertile	3.6 (1.5−8.6)	3.0 (1.2−7.2)	2.8 (1.1−7.3)	2.9 (1.7−4.9)	2.5 (1.4−4.2)	2.1 (1.1−3.7)
P for trend	0.002	0.007	0.04	<0.001	0.001	0.02
Log DcR3
1-SD unit increase	1.7 (1.2−2.6)	1.6 (1.2−2.4)	1.4 (1.1−2.1)	1.7 (1.3−2.1)	1.6 (1.2−2.0)	1.3 (1.1−1.7)
P value	0.007	0.03	<0.05	<0.001	0.001	0.04

Hazard ratios and 95% confidence intervals were derived from Cox regression analysis with DcR3 taken into account as a time-dependent covariate. The multivariate model included variables for age, sex, smoking status, diabetes, prior cardiovascular disease, body mass index, total cholesterol, systolic BP, hemodialysis duration, urea Kt/V, serum albumin, hemoglobin, IL-6, and vascular cell adhesion molecule-1. DcR3, decoy receptor 3.

View popup

Table 5.

Prediction gain by DcR3 over a model of traditional cardiovascular risk factors

Multivariate Cox Hazards Model	HR (95% CI)	P Value	AIC^a	ΔAIC^b
Cardiovascular mortality
cardiovascular risk factors			335.5	0.0
+ VCAM-1 (1 SD unit)	1.8 (0.7−4.4)	0.15	346.7	11.2
+ albumin (1 g/L)	0.9 (0.6−1.3)	0.48	345.0	9.5
+ hs-CRP (1 SD unit)	1.6 (0.9−2.5)	0.05	341.4	5.9
+ IL-6 (1 SD unit)	1.7 (1.1−2.7)	0.01	339.3	3.8
+ DcR3 (1 SD unit)	1.5 (1.1−2.4)	0.03	341.5	6.0
All-cause mortality
cardiovascular risk factors			872.5	0.0
+ VCAM-1 (1 SD unit)	1.7 (1.0−3.1)	0.03	889.6	17.1
+ albumin (1 g/dl)	0.7 (0.5−0.9)	0.01	882.7	10.2
+ hs-CRP (1 SD unit)	1.3 (0.9−1.7)	0.05	892.3	19.8
+ IL-6 (1 SD unit)	1.6 (1.2−2.1)	0.01	885.4	12.9
+ DcR3 (1 SD unit)	1.5 (1.2−2.0)	<0.01	884.0	11.5

DcR3, decoy receptor 3; HR, hazard ratio; 95% CI, 95% confidence interval; AIC, Akaike Information Criterion; VCAM-1, vascular cell adhesion molecule-1; hs-CRP, high-sensitive C-reactive protein.
↵a AIC (15) for a given model is a function of its maximized log-likelihood and the number of estimable parameters (K): AIC = −2 (maximized log-likelihood) + 2K. In the baseline Cox proportional hazards model, there were six traditional cardiovascular risk factors, including age, sex, smoking status, diabetes, total cholesterol, and systolic BP. VCAM-1, albumin, hs-CRP, IL-6, or DcR3 was added subsequently in order to assess the independent predictive contribution by DcR3.
↵b ΔAIC is the difference between the AIC of baseline model and the AIC of VCAM-1, albumin, hs-CRP, IL-6, or DcR3, respectively. Therefore, the lower the ΔAIC, the higher prediction gain of pertinent biomarker.