Pathogenic Significance of Hypertrophic Mesothelial Cells in Peritoneal Effluent and Ex Vivo Culture

Advances in Peritoneal Dialysis, Vol. 20, 2004 M. Auxiliadora Bajo, Gloria del Peso, M. Angeles Castro, Antonio Cirugeda, 1 M. Jose Castro, Teresa Olea, Olga Costero, José A. Sánchez Tomero, 1 Cándido Díaz, Rafael Selgas 1 Pathogenic Significance of Hypertrophic Mesothelial Cells in Peritoneal Effluent and Ex Vivo Culture The presence of hypertrophic mesothelial cells (HMCs) in peritoneal effluent (PE) has been considered a possible marker for peritoneal sclerosis. We conducted the present study to evaluate if the presence of HMCs in PE or in culture was related to peritoneal function alterations or to the development of sclerosing peritonitis. We prospectively studied 32 new peritoneal dialysis (PD) patients every 4 months, determining the presence of HMCs in culture (completing 129 studies in total). We isolated mesothelial cells from nocturnal PE and cultured them ex vivo in T-25 flasks. Cell morphology was estimated using the May Grünwald/Giemsa method. We also examined the histories of a large patient group to determine HMCs directly in PE, and we evaluated 4 of those patients (6%) who showed persistent HMCs. In 10 of 32 prospectively studied patients, we found HMCs during the culture phase. The cells appeared in the first evaluation in 4 patients and in subsequent cultures in the remaining 6 patients. Ultrafiltration (UF) and solute transport capacity in the 10 patients were similar to those of patients who did not show HMCs. Demographic parameters were not different between the two groups. None of the prospectively studied patients showed any clinical or peritoneal functional abnormality during the study. Cultures performed after the observation of HMCs showed very poor growth capacity. The evolution of the 4 patients in the historic group occurred as follows: 1 patient transferred to hemodialysis 2 years after the observation of HMCs. 1 patient died of an unrelated cause after 1 year on PD. From: Hospitales Universitarios La Paz and 1 La Princesa; Grupo de Estudios Peritoneales de Madrid; Instituto Reina Sofía de Investigaciones Nefrológicas; and Fundación Renal Iñigo Alvarez de Toledo, Madrid, Spain. 1 patient received a successful kidney graft 5 years after the observation of HMCs. 1 patient developed type I UF failure 10 years after the first observation. We conclude that the presence of HMCs is not necessarily a marker of sclerosing peritonitis and that such cells may be observed in cultures from new PD patients with no peritoneal function abnormalities. Key words Mesothelial cells, mesothelial hypertrophy Introduction Early diagnosis of peritoneal abnormalities is necessary to avoid irreversible consequences in patients on peritoneal dialysis (PD). For many physiologic processes, body-fluid cytology is a recognized tool for less invasive or even noninvasive diagnosis. Cytology obtained from PD effluent has demonstrated its usefulness in representing membrane status (1). Furthermore, mesothelial cells from peritoneal effluent (PE) have demonstrated sufficient growth capacity ex vivo to reproduce the cell features of peritoneal tissue (2). Although the presence of hypertrophic mesothelial cells (HMCs) in PE has been considered a marker for peritoneal sclerosis (3,4), other authors have found no remarkable functional abnormalities suggestive of severe peritoneal damage (5). We conducted the present study to evaluate whether the presence of HMCs in PE or in culture was related to peritoneal function alterations or to the development of sclerosing peritonitis. Patients and methods Patients We prospectively studied 32 new PD patients (17 men, 15 women; mean age 48.1 ± 17.6 years) to determine

44 the presence of HMCs in culture. All patients were treated with standard PD solutions. Every 4 months, patients were asked to collect nocturnal PE (129 studies completed in total). Peritoneal function was assessed by calculating mass transfer coefficients (MTCs) for urea and creatinine, using a previously described methodology (6,7). Ultrafiltration (UF) capacity was measured using a 3.86% glucose dialysis solution during a 4-hour dwell. Mesothelial cell culture Mesothelial cells were isolated from nocturnal PE and cultured ex vivo. Bags were drained at the dialysis center and immediately processed for mesothelial cell isolation. The methodology for mesothelial cell culture has been previously described (8 10). The number of mesothelial cells was estimated in a Neubauer chamber. Afterward, cells were seeded into 25 cm 2 tissue-culture flasks and counted at confluence. For mesothelial cell subculture, cells were resuspended and seeded into 24-well plates at 2 10 4 cells per well. Viability of the mesothelial cells was assessed by trypan blue exclusion. Cells were identified as mesothelial cells at isolation from the PE and after trypsinization from confluent 25 cm 2 flasks. Cell morphology was assessed in cytospin optical microscopy slides [Cytospin (500 rpm): Thermo Electron Corporation, Woburn, MA, U.S.A.] stained with May Grünwald/Giemsa. The HMCs were defined by size (>100 µm in diameter). Cells were further characterized by immunohistochemical staining. Mesothelial cells were identified by positive staining for cytokeratin and negative staining for CD45, HLA DR, and von Willebrand factor. All antibodies were supplied by Dako Corporation (Glostrup, Denmark). Results Bajo et al. Studies with cultures of mesothelial cells obtained ex vivo In 10 of 32 patients (31.3%), HMCs were found during the culture phase. All were observed in T25 flask cultures [Figure 1]. In 4 patients, the cells appeared only in the first evaluation; in the remaining 6 patients, the cells appeared in subsequent cultures. In 1 patient, HMCs were detected twice. Cultures performed after observation of HMCs showed very poor growth capacity (Figure 2). We found no significant differences in age, sex, or cause of renal failure between the patients with and without HMCs. All of the patients showed no clinical or peritoneal functional abnormalities during the study. FIGURE 1 Morphology of hypertrophic mesothelial cells in culture (T25 flask cultures). Studies with HMCs obtained from PE (in vivo) We also evaluated 4 other patients who persistently showed HMCs in vivo at long term in a prior study. Those patients repeatedly showed large, flat mesothelial cells with a diameter greater than 100 µm (5). Statistical analysis Results are presented as mean ± standard deviation. The Mann Whitney test for nonparametric data was used. Probabilities less than 0.05 were considered statistically significant. FIGURE 2 Cultures performed after the observation of hypertrophic mesothelial cells.

Hypertrophic Mesothelial Cells in PD 45 Urea and creatinine MTCs were not different in either group (with HMCs: 22.1 ± 4.5 ml/min and 10.8 ± 3.1 ml/min respectively; without HMCs: 21.8 ± 4.2 ml/min and 10.5 ± 3.8 ml/min respectively). Ultrafiltration capacity was similar in both groups of patients (with HMCs: 875 ± 357 ml; without HMCs: 900 ± 393 ml). No patient showed clinical or functional data compatible with peritoneal sclerosis. Studies with HMCs obtained in vivo Table I shows the evolution of patients who had persistent HMCs in vivo. Two years after the observation of HCMs, 1 patient was transferred to hemodialysis because of peritonitis. One patient died on PD from an unrelated cause after 1 year. Another patient received a successful kidney graft 5 years later, and the final patient was transferred to hemodialysis because of a type I UF failure. Figure 3 shows representative HMCs in PE. FIGURE 3 Morphology of hypertrophic mesothelial cells in peritoneal effluent. Discussion Mesothelial transformation as a consequence of PD has previously been documented by our group (2). In the course of practicing ex vivo cultures of mesothelial cells, we have occasionally observed hypertrophy in the flasks, a phenomenon that has unknown significance. Some years ago, during our first studies assessing PE cytology, we detected and described HMCs in 4 patients (5). Although both observations represent a similar transformation by mesothelial cells, they happened under different conditions. We have therefore attempted to determine possible relationships. Some authors have mentioned the presence of giant mesothelial cells with the development of peritoneal sclerosing syndromes (4). Peritoneal sclerosis has been reported as a frequent complication in patients on longterm PD (11 13). The condition may progress to sclerosing encapsulating peritonitis, which is a serious complication of the PD modality (12,13). Because we have had the opportunity to observe some patients who showed HMCs in vivo at long-term, we tried in the present study to reply to the questions posed by authors who have associated HMCs with peritoneal sclerosis (3,4). Our results fail to demonstrate that ex vivo HMCs that is, in HMCs culture are related to demography or peritoneal function. However, our most interesting finding is related to the loss of subsequent growth capacity after the first detection of HMCs in culture. Cell hypertrophy could be the consequence of an arrest of the cell cycle. Mesothelial cells usually grow in culture by duplicating themselves until they reach confluence (8 10). When the growth capacity of the cells is limited or exhausted, confluence is delayed or fails to occur. That is the situation in which HMCs are more TABLE I Follow-up of patients who present hypertrophic mesothelial cells in peritoneal effluent Patients 1 2 3 4 Time on PD at diagnosis (years) 3 11 2 6 Age (years) 26 38 65 35 Sex Female Female Male Male Cause of ESRD CGN CGN Diabetes Diabetes Time of follow-up (years) 10 5 1 2 Cause of PD drop-out Membrane failure Renal allograft Death Peritonitis PD = peritoneal dialysis; ESRD = end-stage renal disease; CGN = chronic glomerulonephritis.

46 frequently observed. The coincidence of that phenomenon with the lack of mesothelial cell growth capacity in later assays and the reproducibility of peritoneal tissue findings by effluent cytology (2) are all very suggestive of the loss of mesothelial cell growth capacity in vivo. Regarding HMCs found in vivo in PE, our findings do not support the presence of coexisting sclerosing syndromes. The fact that the patients continued on uneventful PD (up to 10 years after the first detection) demonstrates that sclerosing peritoneal syndrome was definitely not present at the time of culture. We therefore believe that the proposal of other authors (4) to suspend PD when HMCs are observed so as to prevent the development of sclerosing syndromes is inappropriate. The change of mesothelial cells into a hypertrophic morphology is a complex phenomenon that requires more study. We think that the detection of HMCs, either in vivo or ex vivo, represents cellcycle arrest probably as a consequence of a negative relationship with dialysate components. Two different approaches based on mesothelial cell biocompatibility and cellular rest, which combine several weeks of peritoneal rest (14) with restarting PD using more biocompatible fluids, would likely be a more appropriate treatment. The possibility of continuous monitoring for the presence of HMCs in effluent makes the latter approach possible. Conclusion The presence of HMCs is not necessarily a marker of sclerosing peritonitis. Such cells may be observed in culture from new PD patients with no peritoneal functional abnormalities. References 1 Fernández de Castro M, Selgas R, Jiménez C, et al. Cell populations present in the nocturnal peritoneal effluent of patients on continuous ambulatory peritoneal dialysis and their relationship with peritoneal function and incidence of peritonitis. Perit Dial Int 1994; 14:265 70. 2 Yañez Mó M, Lara Pezzi E, Selgas R, et al. Peritoneal dialysis and epithelial-to-mesenchymal transition of mesothelial cells. N Engl J Med 2003; 348: 403 13. 3 Yamamoto T, Izumotani T, Otoshi T, Kim M. Morphological studies of mesothelial cell in CAPD effluent and their clinical significance. Am J Kidney Dis 1998; 32:946 52. Bajo et al. 4 Izumotani T, Ishimura E, Yamamoto T, et al. Correlation between peritoneal mesothelial cell cytology and peritoneal histopathology with respect to prognosis in patients in continuous ambulatory peritoneal dialysis. Nephron 2001; 89:43 9. 5 Selgas R, Fernández de Castro M, Viguer JM, et al. Transformed mesothelial cells in patients on CAPD for medium to long-term periods. Perit Dial Int 1995; 15:305 11. 6 Selgas R, Fernandez Reyes MJ, Bosque E, et al. Functional longevity of the human peritoneum: how long is continuous peritoneal dialysis possible? Results of a prospective medium long-term study. Am J Kidney Dis 1994; 23:64 73. 7 Selgas R, Bajo MA, Paiva A, et al. Stability of the peritoneal membrane in long-term peritoneal dialysis patients. Adv Ren Replace Ther 1998; 5:168 78. 8 Diaz C, Selgas R, Castro MA, et al. Ex vivo proliferation of mesothelial cells directly obtained from peritoneal effluent: its relationship with peritoneal antecedents and functional parameters. Adv Perit Dial 1998; 14:19 24. 9 Castro MA, Diaz C, Bajo MA, et al. Methods to assess the ex vivo growth capacity of mesothelial cells obtained directly from peritoneal effluent [Spanish]. Nefrologia 2000; 20:277 83. 10 Bajo MA, Selgas R, Castro MA, et al. Icodextrin effluent leads to a greater proliferation than glucose of human mesothelial cell studied ex vivo. Perit Dial Int 2000; 20:742 7. 11 Mateijsen MA, van der Wal AC, Hendriks PM, et al. Vascular and interstitial changes in the peritoneum of CAPD patients with peritoneal sclerosis. Perit Dial Int 1999; 19:517 25. 12 Rigby RJ, Hawley CM. Sclerosing peritonitis: the experience in Australia. Nephrol Dial Transplant 1998; 13:154 9. 13 Hendriks PM, Ho-dac-Pannekeet MM, ban Gullik TM, et al. Peritoneal sclerosis in chronic peritoneal dialysis patients: analysis of clinical presentation, risk factors, and peritoneal transport kinetics. Perit Dial Int 1997; 17:136 43. 14 De Alvaro F, Castro MJ, Dapena F, et al. Peritoneal resting is beneficial in peritoneal hyperpermeability and ultrafiltration failure. Adv Perit Dial 1993; 9:56 61. Corresponding author: M. Auxiliadora Bajo, MD, Servicio de Nefrología, Hospital Universitario La Paz, Paseo Castellana 261, Madrid 28046 Spain. E-mail: mabajo.hulp@salud.madrid.org

Advances in Peritoneal Dialysis, Vol. 20, 2004 Teresa Grzelak, Beata Szary, Krystyna Czyzewska Effect of Icodextrin on Transperitoneal Uric Acid and Albumin Transport In Vitro In the present study, we compared the influence of icodextrin (2 g/dl) on uric acid (20 mg/dl) and albumin (1 g/dl) transfer through isolated rabbit parietal peritoneum. In separate series of experiments, we used a mathematical model to calculate the transport rate of solutes from the interstitial to the mesothelial side of peritoneal membrane (I M) and in the opposite direction (M I) before and after introduction of glucose polymer. In control conditions, the rates of uric acid and albumin passage across the rabbit peritoneum remained constant. Mean values [x ± standard error of the mean (SEM)] of the diffusive permeability coefficient P were 1.936 ± 0.324 (I M) and 2.078 ± 0.186 (M I) for uric acid and 0.341 ± 0.066 (I M) and 0.389 ± 0.084 (M I) for albumin (all 10 4 cm/s), respectively. The introduction of glucose polymer into the experimental system did not alter the I M transport of either solute; but, in the opposite direction, it caused a 17% decline in the P for uric acid (p < 0.05), and an increase of 77% in the same parameter for albumin (p < 0.03). Hence, I M transfer dominated M I transfer of uric acid (p < 0.04). In contrast, in albumin transport, M I was higher than I M (p < 0.02). We conclude that, in vitro, icodextrin modifies the diffusive permeability of the peritoneum and induces transport asymmetry for some small and large solutes. Key words Peritoneal transport, icodextrin, uric acid, albumin From: Department of Chemistry and Clinical Biochemistry, Poznan University of Medical Sciences, Poznan, Poland. Introduction Glucose polymer (called icodextrin) has been used in dialysis fluid to respond to the disadvantages of longterm glucose use during peritoneal dialysis (PD), especially in patients with a highly permeable peritoneum (1 3). Glucose polymer has been also proposed as a carrier solution in intraperitoneal chemotherapy, keeping higher doses of drugs in the peritoneal space and reducing the concentration of those drugs in the systemic circulation (4). Recent analyses also demonstrate the possibility that icodextrin can assist in preventing adhesion formation caused by intraperitoneal pharmacotherapy (5). Application of glucose polymer during PD causes both positive and negative effects. It is observed to produce effective and sustained ultrafiltration, with minor absorption to the systemic circulation and lower osmotic stress. It also lessens the patient s caloric load, so that hyperglycemia and hyperinsulinemia are absent, with the concomitant achievement of a more favorable lipid and lipoprotein profile than that produced by glucose-based solutions. Disadvantageous changes include accumulation of breakdown products, a small decline in serum sodium and chloride, cytotoxic effects on monocytes and cultured mesothelium, chemical peritonitis, and DNA injury through a mechanism of lipid peroxidation. Icodextrin also modifies the transport functions of the peritoneum (1 3,6,7). Peritoneal transfer characteristics with the use of icodextrin-based fluid have previously been investigated in both humans and animals, with conflicting results. Also, the studies looked at solute transport in vivo (1,2,7). Our aim in the present study was therefore to examine, in vitro, the direct effect of glucose polymer on the bidirectional transport rate for small (uric acid) and large (albumin) molecules across the peritoneal membrane. Material and methods Fragments of parietal peritoneum from New Zealand male rabbits were placed into a modified Ussing-chamber system as an experimental model. The tissue (with an active membrane surface area amounting to

48 1.1 cm 2 ) was connected through a peristaltic pump to a fluid reservoir containing Hanks solution, which circulated at the rate of 11 ml/min. Adequate oxygen content in the medium and a constant ph (7.4) were both maintained by continuous bubbling of a gas mixture consisting of 5% CO 2 and 95% O 2. The entire system was maintained at 37 C during the experiments (8). The transfer rates of uric acid (initial concentration gradient: 20 mg/dl; Ubichem, Eastleigh, England) and of albumin (initial concentration gradient: 1 g/dl; Sigma Chemical, St. Louis, MO, U.S.A.) from the interstitial to the mesothelial side (I M) of the peritoneal membrane, and in the opposite direction (M I), were studied in two separate series: Control conditions without icodextrin (120 minutes; n = 22) Before (15 60 min) application of icodextrin, and after (75 130 min for uric acid; 75 120 min for albumin) application of icodextrin [2 g/dl (average molecular weight 14.6 kda); ML Laboratories, Liverpool, England] on the mesothelial side of the peritoneum Sampling of the medium was carried out at regular intervals (10 minutes or 15 minutes). The uric acid concentration was determined using an enzymatic test with uricase (Cormay, Lublin, Poland). For albumin, the Peterson modification of the micro Lowry method (Sigma Chemical, St. Louis, MO, U.S.A.) was applied. A mathematical model of mass transport was used to estimate a diffusive permeability coefficient P (scaled to the surface area of the membrane) for the examined specimens. Changes of P attributable to experimental modifications were determined as percentages of the control value for each individual experiment (that is, separately for each sampler of peritoneal membrane) and are presented as mean value ± standard error of the mean (SEM) for the whole series. In this way, the initial part of an experiment with a particular membrane sample served as a control for the second part. For statistical analysis, the Student t-test for paired and unpaired data was used (8). Results In control conditions, the rates of uric acid and albumin transport from the interstitial to the mesothelial Grzelak et al. side of membrane and in the opposite direction remained constant. For the cases of I M and M I, respectively, the values of the diffusive permeability constant P were 1.936 ± 0.324 and 2.078 ± 0.186 for uric acid and 0.341 ± 0.066 and 0.389 ± 0.084 for albumin (all 10 4 cm/s). Icodextrin introduction at 60 minutes of the experiment did not change I M transport of the analyzed solutes. However, in the opposite direction, glucose polymer caused a decrease of about 17% in uric acid diffusion and an increase of about 77% in albumin transfer. Hence, uric acid and albumin transport showed asymmetry (Figure 1): I M passage dominated M I passage of uric acid. In contrast, in albumin transfer, M I was higher than I M. Discussion The results of previous in vivo studies concerning the effects of icodextrin on transperitoneal transport are ambiguous (1,2,7). Application of 7.5 g/dl glucose polymer instead 3.86 g/dl glucose in stable continu- FIGURE 1 Changes in the diffusive permeability (P) of rabbit peritoneum for uric acid and albumin, expressed as a percentage of the control value ± standard error of the mean obtained before the application of glucose polymer, which was introduced on the mesothelial side of the membrane at 60 minutes of the experiment. I = interstitial side of the peritoneal membrane; M = mesothelial side of the peritoneal membrane.

Icodextrin and In Vitro Transport of Uric Acid and Albumin 49 ous ambulatory peritoneal dialysis (CAPD) patients during a 4-hour dwell did not affect the mass transfer area coefficients (MTACs) of such low molecular weight solutes as urea, creatinine, and urate (1). Similarly, in studies with an animal model in vivo, the urea dialysate-to-plasma (D/P) concentration ratio was not significantly different between two groups of rats injected daily for 30 days with 7.5 g/dl icodextrin or 3.86 g/dl glucose (7). However, urea Kt/V and creatinine clearance values were both higher for icodextrin solution (7.5 g/dl) than for glucose solution (2.27 g/dl) in automated peritoneal dialysis patients (2). Additionally, 7.5 g/dl icodextrin based fluid did not change clearances of albumin and other serum proteins (immunoglobulin, macroglobulin) in CAPD patients, but did increase β 2 -microglobulin clearance in comparison with 1.36 g/dl and 3.86 g/dl glucose solutions [about 26% and 44%, respectively (1)]. In contrast to those studies, Frajewicki and coauthors detected a significant increase in the excretion of protein (including albumin), unrelated to molecular weight, with a single use of glucose polymer (7.5 g/dl) in rats. The effect was more prominently after long-term administration of icodextrin. Notably, the simultaneous use in the animals of a mixture of 3.86 g/dl glucose and 7.5 g/dl icodextrin induced the highest levels of protein in the peritoneal effluent (7). These contrasting results are probably related to interspecies heterogeneity in regard to both the structure and the function of peritoneal membrane and the models used (1,7,9 11). In the present in vitro study, application of icodextrin (2 g/dl) modified diffusive peritoneal permeability and caused asymmetry in the transperitoneal transfer of uric acid and albumin, which, in the control conditions, had been similar in both directions and dependent on molecular weight. The modifications varied for examined compounds. In the case of uric acid, I M diffusion dominated M I diffusion. In contrast, albumin I M transport was lower than that in the opposite direction. The reasons for the observed changes are not known, but can probably be attributed to the separate mechanisms of uric acid and albumin peritoneal transport. Uric acid is a low molecular weight compound, and diffusion is the most important process of its passage across the peritoneal membrane. Among other factors, it depends on the transperitoneal concentration gradient, the molecular weight of molecule, and the properties of the peritoneum (9,11). The mechanism of large-solute peritoneal transport is not clear. Controversy exists concerning whether in vivo transfer of macromolecules (for example, albumin) through the large-pore system occurs mainly by diffusion or by convection induced by a hydrostatic and osmotic pressure gradient. The size, shape, and electric charge of high molecular weight solutes are factors with a potential influence on transport from the plasma to the peritoneal cavity. However, the disappearance of large molecules from the intraperitoneal space seems to be independent of the molecular size (9). In vitro, the role of transmesothelial pinocytotic transport cannot be excluded (11). Furthermore, modifications of transport processes after introduction of glucose polymer need also to consider the physicochemical nature of the icodextrin molecule. That is, the molecule s oncotic properties, high viscosity, and possible degradation to lower molecular weight fragments may result in an increase of fluid osmolality and metabolic activity with regard to peritoneal transport functions (3,6,10). Notably, the asymmetry of uric acid and albumin transperitoneal transport induced by icodextrin (2 g/dl) in vitro may be also observed in vivo during PD. That asymmetry seems to be important and beneficial from the clinical viewpoint. The phenomenon indicates better excretion of certain small uremic toxins from the blood to the peritoneal fluid and lower losses of serum proteins. Conclusion The results of the present experiment show that, in vitro, glucose polymer modifies the diffusive permeability of the peritoneum and induces asymmetry of uric acid and albumin transport an asymmetry that could be favorable for the efficiency of PD. The causes of these effects are not yet recognized. Further studies are needed on how the application of icodextrin solution affects the mechanisms that underlie transport through the peritoneum. Acknowledgment This work was supported in part by the State Committee for Scientific Research, grant number 3 P05A 122 22, 2002 2004, Pathophysiology of the macromolecules transperitoneal transport.

50 Grzelak et al. References 1 Krediet RT, Douma CE, Ho-dac-Pannekeet MM, et al. Impact of different dialysis solutions on solute and water transport. Perit Dial Int 1997; 17(Suppl 2): S17 26. 2 Woodrow G, Stables G, Oldroyd B, Gibson J, Turney JH, Brownjohn AM. Comparison of icodextrin and glucose solutions for the daytime dwell in automated peritoneal dialysis. Nephrol Dial Transplant 1999; 14:1530 5. 3 Cooker LA, Holmes CJ, Hoff CM. Biocompatibility of icodextrin. Kidney Int 2002; 62(Suppl 81): S34 45. 4 Hosie KB, Kerr DJ, Gilbert JA, et al. A pilot study of adjuvant intraperitoneal 5-fluorouracil using 4% icodextrin as a novel carrier solution. Eur J Surg Oncol 2003; 29:254 60. 5 Conroy SE, Baines L, Rodgers K, Deviren F, Verco SJ. Prevention of chemotherapy-induced intraperitoneal adhesion formation in rats by icodextrin at a range of concentrations. Gynecol Oncol 2003; 88: 304 8. 6 Gokal R, Moberly J, Lindholm B, Mujais S. Metabolic and laboratory effects of icodextrin. Kidney Int 2002; 62(Suppl 81):S62 71. 7 Frajewicki V, Kushnir D, Wajsbrot V, Kohan R, Shostak A, Gotloib L. Peritoneal transport after longterm exposure to icodextrin in rats. Nephron 2002; 92:174 82. 8 Czyzewska K, Szary B, Waniewski J. Transperitoneal transport of glucose in vitro. Artif Organs 2000; 24:857 63. 9 Krediet RT. Peritoneal physiology impact on solute and fluid clearance. Adv Ren Replace Ther 2000; 7:271 9. 10 Wang T, Cheng HH, Heimbürger O, Waniewski J, Bergström J, Lindholm B. Effect of peritonitis on peritoneal transport characteristics: glucose solution versus polyglucose solution. Kidney Int 2000; 57: 1704 12. 11 Dobbie JW. Ultrastructure and pathology of the peritoneum in peritoneal dialysis. In: Gokal R, Nolph KD, eds. The textbook of peritoneal dialysis. 1st ed. Dordrecht: Kluwer Academic Publishers; 1994: 17 44. Corresponding author: Teresa Grzelak, Department of Chemistry and Clinical Biochemistry, Poznan University of Medical Sciences, Swiecickiego 6 Str., Poznan 60 781 Poland. E-mail: tgrzelak@amp.edu.pl