Seiji Yamagami

Detection of Peptidoglycan and Endotoxin in Dialysate, Using Silkworm Larvae Plasma and Limulus Amebocyte Lysate Methods

Silkworm larvae plasma (SLP) reagent is activated by peptidoglycan (PG), a fragment of both the gram-positive and gram-negative bacterial cell wall, as well as beta-glucan (BG), a component of fungi. It is possible to measure contamination of gram-positive bacteria quantitatively by combining the conventional limulus amebocyte lysate (LAL) and PG measurement methods. Therefore, a more highly accurate analysis of dialysate can be made using both SLP and LAL methods to detect endotoxin (ET) and/or PG contamination. We studied the effects of contaminated dialysate on human peripheral blood mononuclear cells (PBMC) by producing various cytokines in vitro. Muramyl dipeptide (MDP) was used as the biologically active minimum constituent of PG. A total of 54 dialysate samples were obtained under sterile conditions from 4 sites: (1) reverse osmosis water unit; (2) proportioning unit; (3) multiple dialysate preparation console, and (4) personal dialysate preparation console, at 9 dialysis facilities. To detect bacterial contamination, the samples were measured with LAL(C), LAL(G) and SLP methods. PBMC were collected from 10 healthy controls and from 10 hemodialysis patients and cultured for 24 h with ET, MDP, ET + MDP and contaminated dialysate. IL-1 receptor antagonist (IL-1Ra), IL-1 beta and TNF-alpha in the culture medium supernatants were measured using the ELISA method. PG was not detected in dialysate from sites 1 or 2. However, dialysate from the inlet of the dialyzer at the bedside monitor of the central supply and personal console showed 4.1 +/- 6.1 ng/ml for site 3 (in 7 of 18 samples) and 3.3 +/- 4.6 ng/ml for site 4 (in 3 of 18 samples). Contamination by PG alone and complex contamination by PG and ET were also detected. Furthermore, IL-1Ra, IL-1 beta and TNF-alpha production by PBMC increased in accordance with the concentrations of MDP. Cytokine production was enhanced 5-10 times more where MDP and ET coexisted than where either MDP or ET existed alone, showing the synergic effects of MDP and ET. Based on these results, there is a high possibility that PG may also be a pyrogen in the dialysate prior to this study. ET had been considered the only pyrogen in dialysate. Therefore, it is essential to recognize the existence of both ET and PG in investigating dialysate contamination.

Serum hepatocyte growth factor levels in patients with chronic renal failure.

The serum levels of hepatocyte growth factor (HGF) were determined in chronic renal failure (CRF) patients. Nondialysis patients with renal insufficiency had significantly higher serum HGF than normal subjects (0.34 +/- 0.10 ng/ml, n = 21 vs. 0.19 +/- 0.05 ng/ml, n = 15; p < 0.001), and the elevated serum HGF correlated with their serum creatinine levels. Hemodialysis (HD) patients treated for 5-10 years showed higher serum HGF than those receiving HD for 1 year or less (0.45 +/- 0.14 ng/ml, n = 8 vs. 0.33 +/- 0.11 ng/ml, n = 9; p < 0.05). Continuous ambulatory peritoneal dialysis patients also showed elevated serum HGF levels comparable to those of HD patients. There was no difference in serum HGF levels in HD patients with or without acquired cystic disease of kidney. Consequently, serum HGF is elevated in CRF, which may be attributed to the increased production of HGF in response to the chronic renal injury, the effect of heparin, or reduced removal of serum HGF in CRF patients.

Trace element concentrations in hair, fingernails and plasma of patients with chronic renal failure on hemodialysis and hemofiltration.

Trace element concentrations in hair and fingernails and plasma of nondialyzed, hemodialyzed and hemofiltered patients with chronic renal failure and healthy volunteers were measured by atomic and flameless atomic absorption spectrophotometry and neutron activation analysis. Plasma aluminum concentrations in all three groups of patients were higher than in the controls. Aluminum levels in plasma and red blood cells were higher in hemofiltered patients than in dialyzed patients. The aluminum concentrations in the hair of both nondialyzed and dialyzed patients was higher than in the controls, while that in the hemofiltered patients it was not. These elevated aluminum concentrations appear to be mainly caused by the use of aluminum-contaminated dialysate. Calcium concentrations in the hair of nondialyzed and dialyzed patients were higher than in the controls. Plasma zinc concentrations in all the patients were lower than the controls, and the concentrations in hemofiltered patients were lower than in the dialyzed patients.

Detection of bacteria in dialysate and its antibody in long-term hemodialysis patients.

The relationship between endotoxin, which was detected in the dialysate, and incidence of its antibody formation in long-term hemodialysis patients was studied. Bacteria in the dialysate were first obtained, identified, and cultured, and then endotoxin was obtained from the bacteria by extraction with phenol. Using the endotoxin obtained and Escherichia coli J-5 as the antigen, the endotoxin antibody was measured in the serum of long-term hemodialysis patients by enzyme-linked immunosorbant assay (ELISA). It was found that the percentage of endotoxin antibody-positive patients was significantly higher in hemodialysis patients when compared with healthy controls. It was also significantly higher in hemodialysis patients using synthetic polymer membrane dialyzers compared with those using cuprophan membrane dialyzers. These results suggest that endotoxin antibody detected in hemodialysis patients was derived from the dialysate, and that the dialysate must be kept bacteria free, especially in patients using high-flux, highly permeable membrane dialyzers.

Correlation between Blood Prostaglandins and Blood Pressure in Chronic Renal Failure

Plasma prostaglandins (PGs; PGE2, PGF2 alpha, 6-keto-PGF 1 alpha and TXB2) and plasma renin activity (PRA) were measured in 94 end-stage renal disease (ESRD) patients including 15 undialyzed, 74 maintenance-hemodialyzed and 5 anephric patients, and in 27 healthy controls. In the healthy controls, 6-keto-PGF 1 alpha inversely correlated with age, while TXB2 and the TXB2/6-keto-PGF 1 alpha ratio correlated with age. In the ESRD patients, 6-keto-PGF 1 alpha showed a tendency to decrease with age, and the TXB2/6-keto-PGF 1 alpha ratio significantly correlated with age. The undialyzed group showed significantly higher blood pressure and TXB2 but significantly lower 6-keto-PGF1 alpha compared to the other groups. In the dialyzed group, PGE2 and 6-keto-PGF1 alpha tended to be lower and TXB2 higher compared to the healthy control group. In the dialyzed group, 6-keto-PGF1 alpha correlated inversely with blood pressure independently from PRA. As for PGF2 alpha, it was higher in the undialyzed, dialyzed and anephric groups than in the healthy control group. These results suggested that PGs were involved in blood pressure abnormalities, and that PGI2 played an important role in controlling blood pressure in ESRD patients as one of the depressor factors.

Comparative Studies between Hemofiltration and Hemodialysis on Urea and Isotope-Labelled Sodium Kinetics

Comparative studies on urea and sodium kinetics were performed between hemodialysis (HD) and hemofiltration (HF). 10 patients on HF were examined and 11 patients on HD were then selected to match the HF group on the basis of age, body weight and height. The urea concentration of serum, cerebrospinal fluid (CSF) and total urea removed were measured during HD and HF. The slope of regression line of the serum urea concentration during the 5 h of treatment was significantly higher for HD, while the total urea removed was comparable between the two. On the contrary, CSF reduction in urea was significantly greater after HF, resulting in a smaller urea concentration gradient between serum and CSF. The sodium kinetics during the procedures were estimated using 22Na. Although the total 22Na removed was significantly larger in HF than in HD and body weight loss was comparable between the two groups, the slope of regression line for 22Na was significantly less with HF whereas total serum Na levels for HF equaled HD. The above results indicate that urea and sodium equilibrated more efficiently between the compartments of body fluid in HF. This resulted in a smaller decrease in osmolality as well as in solute concentration gradients between plasma and CSF.

Acetate Kinetics during Hemodialysis and Hemofiltration

A comparison was made between acetate dynamics in hemodialysis (HD) with acetate dialysate and hemofiltration (HF) using acetate substitution fluid. Acetate kinetics was calculated from removed HCO-3 (Rbc), generated HCO-3 (Gbc) and loaded CH3COO- (Lac) during each treatment. Rbc and Gbc were 99 ± 17 and 137 ± 18 mmol/h in HF, and 114 ± 23 and 128 ± 27 mmol/h in HD, respectively. Lac was 142 ± 25 mmol/h in HF and 162 ± 27 mmol/h in HD. Accordingly, the conversion rate (CR) from acetate to bicarbonate was 98 ± 18% in HF and 79 ± 9% in HD. Although the net load of acetate was comparable, CR was significantly greater in HF, resulting in a significantly smaller increment of the serum acetate level. Therefore, fewer symptoms are seen and better normalization of acidosis can be achieved in HF.

Use of immobilized histidine in assay for endotoxin in patients with liver disease

TheLimulus amebocyte lysate (LAL) test has the disadvantage of being influenced by various inhibitors and activators. We have developed a method for the LAL reaction that involves the specific adsorption and isolation of endotoxin using a membrane filter unit and immobilized histidine; in this present study we used the method to measure endotoxin in the plasma of patients with acute or chronic liver disease. The adsorbed endotoxins are separated from LAL-inhibitors or-activators by the membrane filter unit, and their activity is directly assayed with the LAL reagent in a filter cup without any inhibition or activation. The study population consisted of 23 subjects, 3 with fulminant hepatitis and 20 with cirrhosis (9 with esophageal varices and 11 without). All 3 (100%) of the samples of plasma from patients with fulminant hepatitis were positive for endotoxin, as were the samples of 7 (78%) of the 9 patients with cirrhosis and esophageal varices, and 2 (18%) of the 11 patients with cirrhosis but without such varices. The results suggested that this method appears to be useful for assaying the concentration of endotoxin in patients with fulminant hepatitis or cirrhosis of the liver.

Correlation between Blood Prostaglandins, Plasma Lipids and Atherosclerosis in Dialyzed Patients

The correlations between the degree of atherosclerosis, plasma prostaglandins (PGs) and plasma lipids were examined in maintenance hemodialyzed patients with and without diabetes mellitus. The degree of atherosclerosis was evaluated by pulse wave velocity (PWV) of the aorta. Plasma PGs were radioimmunoassayed. PWV was significantly higher in hemodialyzed patients compared to sex- and age-matched healthy controls. PWV correlated with the plasma thromboxane A2 (TXB2) level and TXB2/6-keto-PGF1 alpha ratio in hemodialyzed patients, without a significant difference between the diabetic and nondiabetic groups. The plasma lipid profile was type IV of the WHO classification in both the diabetic and nondiabetic groups, and PWV did not correlate with these lipid abnormalities. Though not significantly, the decreases in plasma PGE2 and 6-keto-PGF1 alpha and the increase in TXB2 correlated with the degree of type IV hyperlipidemia. The results suggest that plasma PG abnormalities might correlate with the degree of atherosclerosis in hemodialyzed patients.

Relationship between endotoxin antibody levels and portal systemic shunt evaluated by per-rectal portal scintigraphy

Abstract The reasons for the high frequency of endotoxaemia in cirrhosis, whether poor liver function or abnormal portal circulation, are not known. Accurate measurement of endotoxin itself is difficult. Instead, in this study an enzyme‐linked immunosorbent assay was used to measure levels of IgA, IgG and IgM antibodies to endotoxin in patients with chronic liver disease and underlying hepatic viral infection. The relationships between the results and clinical symptoms or the presence of a portal systemic shunt were investigated. The median level of IgA antibodies was not different in patients with chronic hepatitis and those with cirrhosis, and the same was found for IgM, but the median level of IgG antibodies was significantly higher in the patients with cirrhosis. When patients with cirrhosis were grouped by the presence or absence of ascites or hepatocellular carcinoma, no significant difference was observed in any of these antibody levels. However, in cirrhotic patients with varices, the level of IgG antibodies to endotoxin was significantly higher than in patients without varices. For evaluation of the portal systemic shunt, the per‐rectal portal shunt index was calculated. There was a significant correlation (R= 0.431, P < 0.001) between the per‐rectal portal shunt index and the level of IgG antibodies to endotoxin. That is, the degree of abnormality in the portal haemodynamics was correlated with the level of IgG antibodies to endotoxin in patients with liver disease.