Abstract
Objectives
To investigate the degree and the determinants of peritoneal homocysteine (Hcy) clearance and to compare measured Hcy clearance with the Hcy clearance predicted based on molecular weight (MW).
Design
Cross-sectional observational analysis.
Setting
Tertiary care institutional dialysis center.
Patients
Sixty-five stable peritoneal dialysis (PD) patients.
Outcome Measures
Fasting blood and 24-hour pooled dialysate effluents were collected for determination of peritoneal clearances of Hcy (CpHcy), urea (CpUr), and creatinine (CpCr). The dialysate-to-plasma creatinine ratio at 4 hours (D/P Cr 4 h) and levels of red cell folate, B12, ferritin, and C-reactive protein (CRP) were measured concurrently. Observed CpHcy was compared with predicted clearance, based on Hcy plasma protein binding and the relative molecular weights of Hcy, urea, and creatinine.
Results
Plasma concentrations of Hcy averaged 24.6 ± 1.1 μmol/L and were elevated above the upper limit of normal in 59 (91%) patients. The mean dialysate concentration of Hcy was 2.9 ± 0.3 μmol/L, equating to a daily peritoneal elimination of 34.6 ± 3.6 μmol. Observed CpHcy was closely approximated by predicted CpHcy (8.7 ± 0.6 L/week/1.73 m2 vs 9.0 ± 0.3 L/week/1.73 m2 respectively, p = 0.55). Patients maintained on automated PD (n = 5) had a CpHcy similar to that of patients treated with continuous ambulatory peritoneal dialysis (8.9 ± 1.0 L/week/1.73 m2 vs 8.7 ± 0.6 L/week/1.73 m2, p = 0.92). The CpHcy was significantly correlated with C-reactive protein (CRP), D/P creatinine, CpUr, CpCr, and peritoneal protein loss, but not with plasma Hcy, albumin, B12, ferritin, age, dialysis duration, peritonitis episodes, or daily dialysate effluent volume. By multivariate analysis, the only variables that remained significant independent predictors of CpHcy were CRP and D/P Cr 4 h. High and high-average transporters had a higher CpHcy than low and low-average transporters (9.7 ± 0.8 L/week/1.73 m2 vs 7.0 ± 0.7 L/week/1.73 m2, p < 0.05), despite comparably elevated plasma Hcy concentrations [25.2 ± 1.5 μmol/L vs 23.4 ± 1.6 μmol/L, p = nonsignificant (NS)].
Conclusions
Elevated plasma concentrations of Hcy are not efficiently reduced by PD. The relatively low peritoneal clearance of Hcy is largely accounted for by a high degree of plasma protein binding and is significantly influenced by peritoneal membrane permeability.
Keywords
The etiology of hyperhomocysteinemia in patients with end-stage renal failure is not well understood, but contributing mechanisms may include reduced renal Hcy metabolism (9,13-15) and relative or absolute deficiencies of folic acid, vitamin B12, or vitamin B6 (6,9,14-17). A recent study has suggested that the mode of dialysis is also an important determinant of plasma Hcy concentration, with hyperhomocysteinemia being more intense and prevalent in hemodialysis patients than in peritoneal dialysis (PD) patients (6).
Despite the small molecular weight (MW) of Hcy, hemodialysis reduces Hcy concentrations in plasma by only about 30%; return to pre-dialysis levels occurs promptly within 24 hours (18-25). Studies of the dialytic removal of Hcy in PD patients have been much less extensive; two recently published papers suggest relatively low elimination rates (39 – 49 μmol/day) (26,27).
The present study therefore aimed to investigate the degree and the determinants of peritoneal Hcy clearance in a large cohort of PD patients at a single center, and to compare measured clearance of Hcy with predicted clearance of Hcy based on molecular weight.
Patients and Methods
Patients
All patients over the age of 18 years who had been on a stable PD regimen for a period of at least 1 month were enrolled in the study. Exclusion criteria included the use of dialysate solutions containing either amino acids or glucose polymers and recent (within 1 month) peritonitis or hospital admission. All patients received a minimum of 8 L of dialysis fluid per day and were advised to take a folic acid supplement (5 mg daily). The prescribed dietary protein and energy intakes per kilogram body weight were 1.2 – 1.3 g and 30 – 35 kcal respectively, depending on body mass index (BMI), age, and nutritional status. All patients gave informed consent before their participation in the study, which was approved by the Princess Alexandra Hospital Research Ethics Committee.
Biochemical Assays
Fasting blood samples and 24-hour pooled dialysate effluents were collected for determination of peritoneal Hcy, urea, and creatinine clearances. Blood samples intended for Hcy analysis were immediately placed on ice and centrifuged for 20 minutes at 2000g and 4°C. Plasma and dialysate aliquots were stored at –80°C until subsequently assayed for total (free + protein-bound) Hcy concentration by high-performance liquid chromatography. The reference range for the laboratory was 5 – 15 μmol/L and the intra-assay coefficients of variation were less than 5% for both plasma and dialysate. Urea and creatinine concentrations were measured using routine automated methods (Hitachi 747: Boehringer Mannheim Corp., Indianapolis, IN, U.S.A.), and dialysate protein concentrations were measured by commercially available protein assay (DC Protein Assay Kit II: BioRad, Hercules, CA, U.S.A.) using bovine serum albumin as the standard. Red blood cell folate (ACS180 folate chemiluminescence assay: Bayer Diagnostics, Sydney, Australia) and plasma B12 (Centaur B12 chemiluminescence assay: Bayer Diagnostics), ferritin (Centaur ferritin chemiluminescence assay: Bayer Diagnostics), and C-reactive protein (CRP) levels (Beckman Array 360 rate nephelometry: Beckman Instruments, High Wycombe, U.K.) were measured concurrently.
Peritoneal and Renal Small-Solute Clearances
Peritoneal clearance was calculated as the product of the volume of dialysate outflow and the dialysate-to-plasma ratio of the given solute (urea, creatinine, or Hcy). Body surface area (BSA) was calculated according to the DuBois formula (28) and clearances were then normalized to 1.73 m2 BSA. Peritoneal Kt/V was determined from peritoneal urea clearance and was normalized for the distribution volume of urea, as estimated by the Watson nomogram (29). Predicted peritoneal Hcy clearance was estimated by an equation (30) based on the known extent of plasma protein binding of Hcy [80% (14)] and the relative molecular weights of urea (60 Da), creatinine (113 Da), and Hcy (135 Da):
Residual renal function was estimated as the arithmetic mean of urinary creatinine and urea clearance (31).
Peritoneal Equilibration Test
Peritoneal permeability was assessed by the peritoneal equilibration test (PET) according to the method described by Twardowski and colleagues (32). The dialysate-to-plasma creatinine ratio at 4 hours (D/P Cr 4 h) was used to characterize peritoneal transport types: 0.34 – 0.49 = low; 0.50 – 0.65 = low-average; 0.65 – 0.81 = high-average; and 0.82 – 1.03 = high.
Statistical Analysis
Results are expressed as mean ± standard error of mean (SEM) for continuous data and frequencies, and as percentages for categorical data. Statistical comparisons between groups were made using the unpaired t-test. Pearson product correlation analysis was used to assess the association between Hcy clearance and other data variables. To determine the significant independent predictors of peritoneal Hcy clearance, a multivariate analysis (general linear model) was constructed to include age, dialysis duration, dialysate effluent volume, average daily dialysate dextrose concentration, previous peritonitis episodes, plasma albumin, CRP, red blood cell folate, plasma B12, normalized peritoneal creatinine clearance, normalized peritoneal urea clearance, D/P Cr 4 h, residual renal clearance, and dialysate protein loss as the explanatory variables. Analyses were performed using the StatView software package, version 4.5 (Abacus Concepts Inc., Berkeley, CA, U.S.A.). A p value less than 0.05 was considered significant.
Results
Patient Characteristics
The study included 65 patients being treated with PD [60 continuous ambulatory peritoneal dialysis (CAPD) patients and 5 automated peritoneal dialysis (APD) patients] at the Princess Alexandra Hospital. Table 1 presents patient characteristics. The causes of end-stage renal failure included chronic glomerulonephritis (15 cases), diabetic nephropathy (13 cases), analgesic nephropathy (9 cases), autosomal dominant polycystic kidney disease (3 cases), renovascular nephrosclerosis (3 cases), reflux nephropathy (3 cases), hypertensive nephrosclerosis (2 cases), cystic dysplasia (1 case), renal trauma (1 case), and unknown (15 cases).
Characteristics of the Patients Participating in the Study (n = 65)
CAPD = continuous ambulatory peritoneal dialysis; APD = automated peritoneal dialysis; D/P Cr 4 h = dialysate-to-plasma ratio of creatinine at 4 hours; CpCl = normalized peritoneal creatinine clearance.
All figures are mean and standard error of mean (SEM), or ratio.
Despite the fact that most patients had an acceptable total (peritoneal + renal) small-solute clearance, plasma concentrations of Hcy averaged 24.6 ± 1.1 μmol/L and were elevated above the upper limit of normal in 59 patients (91%). Plasma levels of Hcy were not correlated with peritoneal urea clearance (r = -0.03, p= 0.79), peritoneal creatinine clearance (r = -0.08, p= 0.52), residual renal clearance (r= 0.13, p = 0.33), red blood cell folate (r= -0.21, p= 0.14), plasma B12 (r= -0.05, p= 0.73), or CRP (r= -0.23, p = 0.10). The 39 patients who were taking their folate supplements (5 mg daily) tended to have lower plasma levels of Hcy than did those who were not taking supplements at all (22.9 ±1.1 μmol/L vs 27.1 ± 2.2 μmol/L, p = 0.06). The compliant patients also exhibited significantly higher red blood cell folate levels (2026 ± 180 nmol/L vs 1338 ± 220 nmol/L, p < 0.05), but did not otherwise differ from the non compliant patients in terms of recorded demographic, clinical, and biochemical variables.
Peritoneal Homocysteine Clearance
The mean dialysate concentration of Hcy and the daily peritoneal Hcy elimination were 2.9 ± 0.3 μmol/L and 34.6 ±3.6 μmol, respectively. Measured peritoneal clearance of Hcy (8.7 ± 0.6 L/week/1.73 m2) was closely approximated by predicted Hcy clearance (9.0 ±0.3 L/week/1.73 m2, p < 0.01) based on the MW of Hcy (135 Da) and on the measured peritoneal clearances of urea (67.3 ± 1.8 L/week/1.73 m2, MW = 60 Da) and creatinine (52.0 ±1.3 L/week/1.73 m2, MW = 113 Da). Clearances were not appreciably different between patients on APD and on CAPD (8.9 ± 1.0 L/week/1.73 m2 vs 8.7 ± 0.6 L/week/1.73 m2, p = 0.92). Homocysteine clearance was significantly correlated with CRP, D/P Cr 4 h, creatinine clearance, urea clearance, and daily peritoneal protein loss, but not with any of the other parameters studied (Table 2). By multivariate analysis, the only variables that remained significant independent predictors of Hcy clearance were CRP and D/P Cr 4 h (Table 3). Neither of these parameters was significantly correlated with plasma Hcy. Similar results were obtained in an analysis that included only the patients who complied with prescribed folate supplements (D/P Cr 4 h: p = 0.001; CRP: p< 0.001).
Variables Significantly Correlated with Peritoneal Homocysteine Clearance by Univariate Analysis in 65 Peritoneal Dialysis Patients
D/P Cr 4 h = dialysate-to-plasma ratio of creatinine at 4 hours; NS = nonsignificant.
Multivariate Analysis of Predictors of Peritoneal Clearance in 65 Peritoneal Dialysis Patients (Model r2 = 0.60, p < 0.05)
D/P Cr 4 h = dialysate-to-plasma ratio of creatinine at 4 hours; NS = nonsignificant.
Patients with a D/P Cr 4 h of 0.65 or more (high and high-average transporters) had greater peritoneal clearances of Hcy than did those with a D/P Cr 4 h of less than 0.65 (9.7 ± 0.8 L/week/1.73 m2 vs 7.0 ± 0.7 L/week/1.73 m2, p< 0.05). However, plasma levels of Hcy were not significantly different between the groups [25.2 ± 1.5 μmol/L vs 23.4 ± 1.6 μmol/L,p = nonsignificant (NS)].
Discussion
The results of the present study indicate that Hcy is cleared inefficiently by peritoneal dialysis. Clearance of Hcy did not correlate with plasma Hcy concentration, and it was significantly associated with CRP and D/P Cr 4 h only on a multivariate analysis. No association was found between Hcy clearance and red blood cell folate, plasma B12, plasma albumin, dialysate protein loss, residual renal clearance, or peritoneal urea or creatinine clearance.
Although Hcy clearance has been extensively examined in the setting of hemodialysis (18-25), this parameter has received scant attention in PD patients. Vychytil et al (26) recently reported a mean daily peritoneal Hcy elimination in 39 CAPD patients of 38.9 μmol, which is similar to the value of 34.6 μmol observed in the present study. Another investigation of 46 CAPD patients by Ducloux and associates (27) reported an average, non normalized Hcy clearance of 14.3 L/week, which again is comparable with that observed in the current report (8.7 L/week/1.73 m2).
A novel finding of the present study was that the measured Hcy clearance was remarkably comparable with the clearance expected by simple diffusion based on the degree of Hcy plasma protein binding (80%), the molecular weight of Hcy, and the relative measured clearances of urea (67.3 L/week/1.73 m2, MW = 60 Da) and creatinine (52.0 L/week/1.73 m2, MW = 113 Da). This finding suggests that the inefficient dialytic removal of Hcy in PD is primarily accounted for by the high degree of plasma protein binding of Hcy, without necessarily having to invoke additional mechanisms such as intra-peritoneal degradation, or active peritoneal reabsorption of Hcy, or both. The poor dialytic removal of Hcy is also likely to account for the lack of correlation observed between plasma levels of Hcy and dialysis adequacy in both this and other reports (9,27).
The present study also found no association between peritoneal clearance of Hcy and plasma concentration of Hcy. This finding contrasts with that of Vychytil and coworkers (26), who noted a significant, positive, linear association between plasma Hcy and daily peritoneal elimination. The meaning of this disparity in results is not entirely clear, but may be partly explained by the Vychytil group's exclusion of methylenetetrahydrofolate reductase C677T homozygotes, which occur in one sixth of the end-stage renal failure population (9,15,33) and which can significantly influence plasma Hcy concentrations. Moreover, the patients in the Vychytil study received either no folate supplementation (66%) or minimal folate supplementation (33%: 0.18 mg/day), a situation known to suppress plasma levels of Hcy (12,16,17,22,34-36). By comparison, the 58% of our patients who took folate (5 mg daily) demonstrated plasma Hcy concentrations 15% lower than those in patients who did not take supplements.
Clearance of Hcy was found to be significantly influenced only by D/P Cr 4 h and CRP. High and high-average transporters had Hcy clearances 40% higher than those of low and low-average transporters. However, this difference did not translate into lower plasma concentrations of Hcy. The positive correlation between CRP and clearance of Hcy did not appear to be mediated by increased peritoneal permeability (37), as CRP and D/P Cr 4 h were poorly associated. Moreover, although plasma concentrations of Hcy have been shown to be sensitive to acute-phase reactions in patients with normal renal function (38), this sensitivity was not observed in the present study, where CRP was not correlated with plasma Hcy. The increase in clearance of Hcy with an increase in CRP therefore remains unexplained.
Conclusions
The present study of a prevalent PD population at a single center shows that peritoneal clearance of Hcy is significantly associated only with peritoneal permeability and CRP levels, and is not associated with vitamin status, demographic factors, residual renal clearance, or dialysis adequacy and prescription. Moreover, despite the small molecular size of Hcy, its dialytic removal via the peritoneal membrane is inefficient owing to a high degree of plasma protein binding, and this inefficient removal does not appreciably influence plasma Hcy concentrations. Efforts to treat hyperhomocysteinemia in PD patients should therefore concentrate on established treatments that augment Hcy metabolism (such as folate supplementation) rather than on peritoneal elimination.
Footnotes
Acknowledgment
The invaluable assistance of the nursing staff of the Princess Alexandra Hospital Peritoneal Dialysis Unit is gratefully acknowledged.