Patients on peritoneal dialysis (PD) encounter peritoneal functional and structural alterations. It is still unknown whether levels of plasminogen activator inhibitor type 1 (PAI-1), matrix metalloproteinases- (MMP-) 2, and vascular endothelial growth factor (VEGF) exhibit dynamic changes in peritoneal effluents. The aim of the present study was to investigate the longitudinal changes in these biomarkers in PD patients and their association with peritoneal small-solute transfer rate (PSTR). This prospective, single-center cohort study included 70 new PD patients. The presence of PAI-1, MMP-2, and VEGF in peritoneal effluents was measured regularly after PD initiation. The association between those biomarkers and 4-hour effluent:plasma creatinine ratio (PSTR) was analyzed. Longitudinal follow-up showed a tendency for PAI-1 (
Peritoneal dialysis (PD) is one of the renal replacement therapies employed in end-stage renal disease. Most common PD solution is glucose-based in clinical practice. There have been previous reports that peritoneal injury inferred from glucose-based PD solution in long-term dwell could lead to peritoneal functional and structural changes [
Plasminogen activator inhibitor type 1 (PAI-1), which has a molecular weight of approximately 50 kD, is produced by various cell types, including endothelial cells and vascular smooth muscle cells [
The gelatinase, matrix metalloproteinases (MMP)-2 (molecular weight 72 kDa), degrades gelatin, collagen type IV, fibronectin, laminin, proteoglycan, and elastin [
Vascular endothelial growth factor (VEGF) is a broad term referring to five isoforms of homodimeric glycoprotein with high heparin affinity [
We hypothesized that glucose-based PD solutions would affect peritoneal structure and function. These alterations could be expressed in terms of concentrations of a variety of biomarkers obtained from PD effluents. Our study aimed to examine the serial changes in the concentration of biomarkers in PD effluents since PD initiation and the association of these changes with PSTR. Three biomarkers, PAI-1, MMP-2, and VEGF, were selected for the present study.
Adult new PD patients who had commenced PD therapy since 2014 in our PD unit were included in the study. The inclusion criteria were patients with new catheter implant and receiving PD therapy for more than three months, age ≥18 years, and stable clinical condition during observational period. The exclusion criteria were uncompleted clinical information, discontinuation of PD therapy within six months due to death, shifting to hemodialysis, kidney transplantation, or transfer to other hospitals, advanced liver disease, malignancy, and incidence of peritonitis during the study period. All subjects were observed from September 29, 2014, to April 26, 2018.
The informed consent was obtained from individual subjects before study commencement. This study was approved by the Committee on Human Research at the Kaohsiung Chang Gung Memorial Hospital (Document no. 102-5925B), and the study was conducted in accordance with the principles of the Declaration of Helsinki.
Blood parameters from hemogram and biochemistry tests were measured once monthly. Standard PETs were performed at the first month and repeated every six months after PD commencement. Residual renal function (RRF) was calculated as the arithmetic mean of 24 h urea nitrogen and creatinine clearance, which were measured one month after PD commencement and at six-month intervals thereafter. RRF was normalized to body surface area using the Du Bois formula and patient body weight [
Peritoneal effluents were collected at overnight for MMP2, VEGF, and PAI-1 measurements when PET was performed at the first month, sixth month, and twenty-fourth month. All measurements were performed using commercial ELISA kits (quantitative sandwich ELISA, R&D Systems, Minneapolis, MN, USA).
The clinical outcome measure was PSTR which was defined as 4-hour PD dialysate:plasma creatinine (D/P Cr) ratio measured at PET.
The baseline characteristics of individuals were summarized as mean and standard deviation, frequency and percentage, or median and interquartile range, as appropriate. The variation of inflammation markers from baseline to 6 months and that of 4-hour D/P creatinine from baseline to 2 years were estimated using paired
A total of 70 new PD patients were enrolled. The mean age was 56.4 years. Gender distribution was equal. Forty-one patients were treated with continuous ambulatory peritoneal dialysis (CAPD), 21 with ambulatory peritoneal dialysis (APD), and eight with CAPD+APD. There were relatively low percentage (n=7, 10%) in using ACEi or ARB or statin (n=30, 42.9%) in the participants. Peritoneal transport at the baseline was categorized as follows: high transport, 20.3%; high average transport, 33.3%; low average transport, 31.9%; and low transport, 14.5% (Table
Baseline characteristics of peritoneal dialysis patients (n = 70).
Variables | Mean | SD |
---|---|---|
Age (years) | 56.4 | ±13.4 |
PD vintage (months) | 2.1 | ±1.5 |
Gender, Male (n, %) | 35 | 50.0% |
Diabetes mellitus (n, %) | 33 | 47.1% |
ACEi/ARB (n, %) | 7 | 10.0% |
Statin (n, %) | 30 | 42.9% |
Laboratory measurements | ||
hemoglobin(g/dL) | 10.4 | ±1.5 |
albumin (g/dL) | 3.6 | ±0.4 |
fasting glucose(mg/dl) | 106 | 93-198 |
PET measurements | ||
Category | ||
high transport | 14 | 20.3% |
high average transport | 23 | 33.3% |
low average transport | 22 | 31.9% |
low transport | 10 | 14.5% |
Total KtV | 2.1 | ±0.5 |
Residual Renal Ccr (ml/min/1.73m2) | 34.9 | 14.4-53.6 |
Total Ccr (ml/min/1.73m2) | 73.3 | 56.8-86.3 |
nPCR(g/day) | 1.1 | ±0.3 |
Paired test
PD, peritoneal dialysis; ACEi, angiotensin converting enzyme inhibitor; ARB, angiotensin receptor blockade; VEGF, vascular endothelial growth factor; PAI-1, plasminogen activator inhibitor-1; MMP2, matrix metalloproteinases 2; D/P Cr, dialysate: plasma creatinine.
4-hour D/P creatinine had increased significantly from baseline (0.69) to 2-years (0.70) (
Temporary trends of dialysate inflammatory markers (n = 70).
Variables | Mean | SD | |
---|---|---|---|
Dialysate inflammatory markers | |||
VEGF(pg/ml) (Baseline) | 21.6 | 19.1-25.4 | |
VEGF(pg/ml) (6 months) | 22.2 | 19.7-27.5 | |
Paired test | 0.04 | ||
PAI-1(ng/ml) (Baseline) | 0.6 | 0.3-1 | |
PAI-1(ng/ml) (6 months) | 1.1 | 0.8-1.6 | |
Paired test | <0.001 | ||
MMP-2(ng/ml) (Baseline) | 32.0 | 21.3-46.4 | |
MMP-2(ng/ml) (6 months) | 32.9 | 23.3-46.6 | |
Paired test | 0.73 | ||
Outcome | |||
Follow-up interval (year) | 2.24 | 1.96-2.46 | |
4-hour D/P Creatinine (Baseline) | 0.69 | 0.57-0.79 | |
4-hour D/P Creatinine (2 years) | 0.70 | 0.64-0.78 | |
Paired test | 0.02 |
Paired test
PD, peritoneal dialysis; ACEi, angiotensin converting enzyme inhibitor; ARB, angiotensin receptor blockade; VEGF, vascular endothelial growth factor; PAI-1, plasminogen activator inhibitor-1; MMP2, matrix metalloproteinases 2; D/P Cr, dialysate:plasma creatinine; PET, peritoneal equilibration test; Ccr, creatinine clearance rate; nPCR, normalized protein catabolic rate.
Pearson’s correlation test showed significant correlation between PSTR at baseline with PAI-1, MMP-2, and VEGF concentrations at baseline and PSTR after two years with PAI-1, MMP-2, and VEGF concentrations at baseline and MMP-2 levels at six months (Table
Correlation between inflammation markers and 4-hour D/P creatinine at different time-point measurements.
Variables | 4-hour D/P Creatinine | 4-hour D/P Creatinine | ||
---|---|---|---|---|
(Baseline) | (2 years) | |||
| | | | |
VEGF(pg/ml) | ||||
Baseline | 0.36 | 0.002 | 0.26 | 0.029 |
6 months | 0.10 | 0.404 | 0.14 | 0.248 |
PAI-1(ng/ml) | ||||
Baseline | 0.40 | < 0.001 | 0.34 | 0.004 |
6 months | 0.34 | 0.005 | 0.20 | 0.110 |
MMP-2(ng/ml) | ||||
Baseline | 0.64 | < 0.001 | 0.32 | 0.006 |
6 months | 0.59 | < 0.001 | 0.56 | < 0.001 |
In univariate analysis, peritoneal effluent levels of VEGF, PAI-1, and MMP2 at baseline showed significant association with 4-hour D/P Cr at two-year measurement. These correlations were absent in multivariate analysis except for the peritoneal effluent PAI-1 (
Linear regression analysis for association between biomarkers, covariates, and 4-hour D/P creatinine measured at two-year follow-up.
Variables | Univariate | Multivariate | ||
---|---|---|---|---|
Beta | | Beta | | |
VEGF(pg/ml) | ||||
Baseline | 0.26 | 0.029 | -0.09 | 0.604 |
6 months | 0.14 | 0.248 | -0.001 | 0.994 |
PAI-1(ng/ml) | ||||
Baseline | 0.34 | 0.004 | 0.36 | 0.039 |
6 months | 0.2 | 0.11 | -0.28 | 0.103 |
MMP-2(ng/ml) | ||||
Baseline | 0.32 | 0.006 | -0.07 | 0.694 |
6 months | 0.56 | <0.001 | 0.69 | <0.001 |
Covariates | ||||
Age (years) | -0.06 | 0.600 | -0.13 | 0.257 |
Gender, Male (n, %) | 0.06 | 0.613 | -0.06 | 0.611 |
Diabetes mellitus (n, %) | 0.07 | 0.564 | 0.01 | 0.902 |
ACEi/ARB (n, %) | 0.19 | 0.107 | 0.17 | 0.113 |
Statin (n, %) | -0.12 | 0.337 | -0.04 | 0.739 |
Hemoglobin(g/dL) | -0.03 | 0.788 | 0.04 | 0.719 |
Albumin(g/dL) | -0.24 | 0.047 | 0.02 | 0.888 |
Fasting glucose(mg/dl) | 0.19 | 0.128 | 0.06 | 0.611 |
Multivariate model includes VEGF, PAI-1, and MMP2 measurements at two time points and adjusted for covariates including age, gender, DM, ACEi/ARB, statin, hemoglobin, albumin, and glucose.
It is well known that glucose-based PD solutions could provoke peritoneal membrane injury [
The formation of fibrin on peritoneal surface has been associated with the appearance of adhesions in PD patients. It is already known that mesothelial cells exhibit fibrinolytic activity associated with t-PA production. In study with abdominal surgery model, PAI-1 was found throughout the submesothelial tissue when inflammatory reaction occurred [
VEGF is a mediator of neoangiogenesis. In long-term PD therapy with glucose-based PD solution, VEGF is present due to local production in the peritoneal membrane [
Present study is subject to several limitations. First, timing for sampling was different from prior studies. Thus, there is a discrepancy in head-to-head comparison. Second, we did not measure plasma concentrations of selected biomarkers. Therefore, we assumed that these biomarkers were produced locally in peritoneal membrane based on results of previous studies. Third, we did not examine the impact of genetic polymorphism in individual biomarkers on peritoneal transport. Fourth, we did not examine the inflammatory status of individual new PD patients during the study period by inflammatory markers test, such as CRP and IL-6. Therefore, the impact of inflammation in peritoneal cavity upon changes in biomarker concentration in PD effluents cannot be examined. Finally, the sample size was relatively small and may influence the statistical analysis. Nevertheless, the strength in our study was that a longitudinal peritoneal solute transport over two years in new PD patients was recorded and we examined the association with biomarkers’ concentrations in peritoneal effluents during the first six-month PD therapy.
Our study demonstrated that PAI-1 and MMP-2 in PD effluents may be used as biomarkers in new PD patients with respect to relevant peritoneal transport parameter. Further studies with larger study population and varying degrees of peritoneal conditions are required to validate the results of present study.
Peritoneal dialysis
Plasminogen activator inhibitor-1
Matrix metalloproteinases 2
Vascular endothelial growth factor
Peritoneal solute transport rate
Peritoneal equilibration test
Tissue-type plasminogen activator
Residual renal function
Dialysate plasma creatinine
Angiotensin converting enzyme inhibitor
Angiotensin receptor blockade
Continuous ambulatory peritoneal dialysis
Automated peritoneal dialysis
Continuous cyclic peritoneal dialysis
Mass transfer area coefficient.
The raw data used to support the findings of this study are available from the corresponding author upon request.
The authors declare that they have no conflicts of interest.
The authors were grateful for the following research grants for the accomplishment of the present study: Kaohsiung Chang Gung Memorial Hospital at Taiwan (CMRPG8D1151), Tianjin Science and Technology Program (No.15ZXLCSY00020), Research Project of Tianjin Municipal Health Commission on key areas of TCM (2018004), the Science & Technology Development Fund of Tianjin Education Commission for Higher Education (2017KJ157), and Research Project of Tianjin Municipal Health Commission on TCM (2017116). The authors were grateful for nurses in the PD unit for their assistance in collection of relevant clinical data.