Pre-Clinical Biocompatibility Testing of Peritoneal Dialysis Solutions

Pre-clinical biocompatibility testing of peritoneal dialysis (PD) solutions has become an integral part of new solution development. The construction of a pre-clinical screening program for solution biocompatibility should take a hierarchical approach, starting with in vitro cell viability and function assays. The selection of cell types and assay systems for the in vitro studies should be broad enough to permit a balanced interpretation. Whenever possible, animal models are recommended for the next hierarchical level of testing, followed by human ex vivo study designs. Designs of the latter sort provide evidence that a new solution formulation is exerting an altered biological response in vivo; the response is not purely an in vitro artifact or restricted to a given animal species. This article discusses the various approaches available for biocompatibility testing during the pre-clinical phase of solution development, with an emphasis on the advantages and drawbacks of each method.

Early Glycated Albumin, but Not Advanced Glycated Albumin, Methylglyoxal, or 3-Deoxyglucosone Increases the Expression of Pai-1 in Human Peritoneal Mesothelial Cells

Objective

The continuous contact of glucose-containing peritoneal dialysis (PD) fluids with the peritoneum results in the intraperitoneal formation of early and advanced glycation end-products. This nonenzymatic glycation of proteins may cause morphological and functional alterations to the peritoneum, which may contribute to patient dropout from PD therapy. Because fibrinolytic system components have been demonstrated to play an important role in the balance of intraperitoneal generation and degradation of fibrin, we studied the effect of early and advanced glycated human serum albumin, methylglyoxal, and 3-deoxyglucosone on the synthesis of tissue-type plasminogen activator (tPA), as well as its specific inhibitor (PAI-1), in human peritoneal mesothelial cells (HPMC).

Methods

Antigen concentrations in the supernatants of cultured HPMC were measured by ELISA. Northern blot analysis was conducted for mRNA expression. Electrophoretic mobility shift assays were applied to demonstrate the involvement of the transcription factors nuclear factor kappa B (NF-κB) and activator protein-1 (AP-1) in signal transduction.

Results

Incubation of HPMC with early glycated albumin (GHSA) resulted in a time- and concentration-dependent increase in PAI-1 mRNA expression and antigen secretion. In contrast, no changes in PAI-1 synthesis occurred after stimulation with either the 1,2-dicarbonyl compounds methylglyoxal and 3-deoxyglucosone, or with late advanced glycation end-products. tPA synthesis was not affected by any of the tested components. Furthermore, HPMC exposed to GHSA induced NF-κB and AP-1 DNA binding activity, suggesting that GHSA-induced over-expression of PAI-1 is transcriptionally regulated by both transcription factors.

Conclusions

We conclude that Amadori modified glycated albumin upregulates PAI-1 synthesis in HPMC, possibly mediated through the activation of the transcription factors NF-κB and AP-1. The present data support the clinical relevance of the formation of glycated proteins and their involvement in pathological processes in PD patients. Thus, glycated albumin may contribute to an imbalance between intraperitoneal formation and degradation of fibrin that causes peritoneal structural alterations, with subsequent membrane failure.

Glucose Degradation Products in Peritoneal Dialysis Fluids: How they can be Avoided

♦ Objectives

A patient on peritoneal dialysis (PD) uses 3 – 7 tons of PD fluid every year. The result is considerable stress on the peritoneal tissue. Aspects of PD fluids that have been considered responsible for bioincompatibility are low pH, high osmolality, high glucose and lactate concentrations, and the presence of glucose degradation products (GDPs). However, the relative importance of each factor in PD fluid has so far not been investigated. Discovering their relative importance was the aim of the present study.

♦ Methods

Two main methods for investigating biocompatibility were used in this study: cytotoxicity measured as in vitro inhibition of cell growth, and in vitro AGE formation measured as albumin-linked fluorescence.

♦ Results

The two most important factors for determining in vitro bioincompatibility of PD fluids were the presence of GDPs, which caused both severe cytotoxicity and strong AGE promotion, and low pH, which induced severe cytotoxicity.

♦ Conclusions

The biocompatibility of PD fluids can be monitored through fairly simple in vitro methods such as cell proliferation and AGE formation. Bioincompatibility of PD fluids is caused mainly by the presence of GDPs and low pH. These findings correlate well with known clinical bioincompatibility.

Intraperitoneal Infusion of Glucose-Based Dialysate in the Rat—An Animal Model for the Study of Peritoneal Advanced Glycation End-Products Formation and Effect on Peritoneal Transport

Objective

Glucose-based dialysate induces non enzymatic glycation within the peritoneal cavity. To evaluate the relationship between the formation of advanced glycation end-products (AGEs) and peritoneal transfer for small solutes and macromolecules, we developed a model of simulated peritoneal dialysis (PD) in normal rats.

Methods

Male albino rats of the Charles River strain were divided into two sets of 3 groups (15 – 25 rats in each group). In the experimental (E) group, the rats were intra-peritoneally (IP) injected daily with a commercially available 4.25% dextrose solution. In the control puncture (CP) group, the peritoneum was punctured daily, but no PD solution infused. In an age-matched control (CC) group, no intervention was given. Two study protocols were used. Protocol A (duration 20 weeks) consisted of a daily IP injection of 10 mL PD solution per 100 g body weight. In protocol B, a double volume of PD solution was introduced (20 mL per 100 g body weight). At 9, 16, and 20 weeks in protocol A, and at 9 weeks in protocol B, urea, creatinine, microalbumin [(MAL) measured using specific anti-rat albumin monoclonal antibody], and AGEs (measured by fluorescent assay with excitation at 370 nm and emission at 440 nm) were measured in peritoneal effluent and serum.

Results

At no time during the study were AGEs detected in serum from any group in either protocol. In both protocols, no differences were found between the control groups (CP, CC) with respect to all parameters. In protocol A, the dialysate-to-plasma ratio (D/P) of urea was significantly higher in the experimental group as compared with the control groups at 9, 16, and 20 weeks [9 weeks: 0.59 ± 0.03 (E) vs 0.39 ± 0.02 (CP) vs 0.46 ± 0.02 (CC), p < 0.0004 and p < 0.002, respectively; 16 weeks: 0.71 ± 0.08 (E) vs 0.42 ± 0.01 (CP) vs 0.46 ± 0.01 (CC), p < 0.0001 and p < 0.02, respectively; 20 weeks: 0.57 ± 0.03 (E) vs 0.39 ± 0.01 (CP) vs 0.41 ± 0.02 (CC), p < 0.002 and p < 0.004, respectively]. At 16 and 20 weeks, dialysate MAL levels were significantly increased in group E [16 weeks: 354.00 ± 80.35 μg/mL (E) vs 134.75 ± 14.36 μg/mL (CP) vs 110.69 ± 7.83 μg/mL (CC), p < 0.04 and p < 0.03, respectively; 20 weeks: 283.17 ± 14.71 μg/mL (E) vs 105.14 ± 12.11 μg/mL (CP) vs 135.50 ± 19.03 μg/mL (CC), p < 0.00001 and p < 0.0001, respectively]. In protocol B, at completion of the study (week 9), D/P urea, effluent MAL, and AGEs were significantly higher in the experimental group as compared with the control groups [D/P: 0.67 ± 0.04 (E) vs 0.46 ± 0.07 (CP) vs 0.41 ± 0.02 (CC), p < 0.0002 and p < 00001, respectively; MAL: 336.8 ± 63.30 μg/mL (E) vs 125.71 ± 16.77 μg/mL (CP) vs 119.00 ± 39.75 μg/mL (CC), p < 0.008 and p < 0.007, respectively; AGEs: 265.77 ± 33.49 U/mg creatinine (E) vs 163.10 ± 21.99 U/mg creatinine (CP) vs 83.17 ± 22.66 U/mg creatinine (CC), p < 0.02 and p < 0.001, respectively]. Peritoneal effluent AGEs were found to be significantly correlated with D/P urea and dialysate MAL ( r = 0.42, p < 0.04, and r = 0.7, p = 0.00001, respectively).

Conclusions

In situ generation of AGEs constitutes the chief origin of peritoneal AGEs. Advanced glycation end-products affect peritoneal permselectivity for both small and large solutes. The rat model of simulated peritoneal dialysis developed in this experiment provides a reliable method for studying peritoneal AGE formation and effect on peritoneal transfer of small solutes and macro-molecules.

Biological Significance of Reducing Glucose Degradation Products in Peritoneal Dialysis Fluids

Carbohydrates are not stable when exposed to energy; they degrade into new molecules. In peritoneal dialysis (PD) fluids, degradation of glucose occurs during the heat sterilization procedure. The biological consequences of this degradation are side effects such as impaired proliferation and impaired host defense mechanisms, demonstrated in vitro for a great variety of cells.

Several highly toxic compounds—such as formaldehyde and 3-deoxyglucosone—have been identified in PD fluids. Carbonyl compounds, apart from being cytotoxic, are also well-known promoters of irreversible advanced glycation end-products (AGEs), which might participate in the long-term remodeling of the peritoneal membrane.

Various approaches can be used to reduce the formation of glucose degradation products (GDPs) during heat sterilization. Some examples are shortening the sterilization time, lowering the pH, removing catalyzing substances, and increasing glucose concentration. The latter three factors are employed in the multi-compartment bag with a separate chamber containing pure glucose at high concentration and low pH.

Gambrosol trio, a PD fluid produced in this way, shows reduced cytotoxicity, normalized host defense reactions, less AGE formation, and reduced concentrations of formaldehyde and 3-deoxyglucosone. Moreover, in the clinical situation, the fluid turns out to be more biocompatible for the patient, causing less mesothelial cell damage, which in the long term could lead to a more intact peritoneal membrane.

Conclusion

Glucose degradation products in heat-sterilized fluids for peritoneal dialysis are cytotoxic, promote AGE formation, and cause negative side effects for the patient. Using improved and well-controlled manufacturing processes, it is possible to produce sterile PD fluids with glucose as the osmotic agent but without the negative side effects related to GDPs.

Degradation in Peritoneal Dialysis Fluids May be Avoided by Using Low pH and High Glucose Concentration

Objective

When glucose is present in a medical fluid, the heat applied during sterilization leads to degradation. The glucose degradation products (GDPs) give rise to bioincompatible reactions in peritoneal dialysis patients. The extent of the degradation depends on a number of factors, such as heating time, temperature, pH, glucose concentration, and catalyzing substances. In the present work, we investigated the influence of pH and concentration in order to determine how to decrease the amounts of GDPs produced.

Design

Glucose solutions (1% - 60% glucose; pH 1 - 8) were heat sterilized at 121°C. Ultraviolet (UV) absorption, aldehydes, pH, and inhibition of cell growth (ICG) were used as measures of degradation.

Results

Glucose degradation was minimum at an initial pH (prior to sterilization) of around 3.5 and at a high concentration of glucose. There was considerable development of acid degradation products during the sterilization process when the initial pH was high. Two different patterns of development of UV-absorbing degradation products were seen: one below pH 3.5, dominated by the formation of 5-hydroxy-methyl-2-furaldehyde (5-HMF); and one above, dominated by degradation products absorbing at 228 nm. 3-Deoxyglucosone (3-DG) concentration and the portion of 228 nm UV absorbance not caused by 5-HMF were found to relate to the in vitro bioincompatibility measured as ICG; there was no relation between 5-HMF or absorbance at 284 nm and bioincompatibility.

Conclusion

In order to minimize the development of bioincompatible GDPs in peritoneal dialysis fluids during heat sterilization, pH should be kept around 3.2 and the concentration of glucose should be high. 5-HMF and 284 nm UV absorbance are not reliable as quality measures. 3-DG and the portion of UV absorbance at 228 nm caused by degradation products other than 5-HMF seem to be reliable indicators of bioincompatibility.

Effects of Neutral pH and Reduced Glucose Degradation Products in a New Peritoneal Dialysis Solution on Morphology of Peritoneal Membrane in Rats

BACKGROUND

In vitro studies have shown that pH and glucose degradation products (GDPs) in the dialysate are determinant factors for the biocompatibility of peritoneal dialysis (PD) treatment. The present study was thus designed to evaluate whether a newly developed PD solution, which features neutral pH levels and a low GDP concentration, influences tissue damage of the peritoneal membrane in an in vivo setting, and which factor is more critical to the histological changes.

METHODS

Rats were injected 3 times per day during 1 or 4 weeks with 10 ml of various PD fluids (group G, acidic pH, high GDPs; group S, neutral pH, low GDPs; or group A, acidic pH, low GDPs). When the experimental period was over, the mesothelial cell monolayers of the animals were taken and studied with population analysis, and peritoneal membranes were obtained from the abdominal wall for immunohistochemical examination with proliferating cell nuclear antigen (PCNA) and for measurement of thickness of the peritoneal specimens.

RESULTS

The density of the mesothelial cell monolayer and the number of fibroblast-like cells in group S were significantly less than in group G at 1 and 4 weeks' injection. PCNA-positive nuclei in group S were significantly less than in group G for only the 1-week injection set (group G, 2.03 +/- 0.95; group S, 0.85 +/- 1.18 nuclei/1 x 10(4) microm2). At 4 weeks, the peritoneal thickness of group S (6.32 +/- 0.53 microm) was significantly less than that of group G (7.94 +/- 0.77 microm), There was no significant difference between groups S and A throughout the whole study period except for the result of the number of fibroblast-like cells.

CONCLUSION

These results indicate that a PD solution with a neutral pH and low GDPs proved more biocompatible with the peritoneal membrane than a solution with an acidic pH and high GDPs. Furthermore, the level of the GDPs has more impact on tissue damage of the peritoneal membrane than the pH in the short term.