Glucose and Mannitol Have Different Effects on Peritoneal Morphology in Chronically Dialyzed Rats

Transcription

1 Advances in Peritoneal Dialysis, Vol. 19, 2003 Arkadiusz Styszynski, Beata Kwiatkowska, Katarzyna Wieczorowska Tobis, Andrzej Breborowicz, Dimitrios G. Oreopoulos 1 Glucose and Mannitol Have Different Effects on Peritoneal Morphology in Chronically Dialyzed Rats In previous studies, we showed that glucose reduces the morphologic changes in rat peritoneum caused by chronic intraperitoneal administration of 0.9% NaCl solution. In the present study, we set out to determine if the observed results were attributable to hyperosmolarity or to the metabolic effect of the glucose. Intraperitoneal catheters were implanted in 19 rats. The animals were then intraperitoneally exposed twice daily for 30 days to 20 ml 0.9% saline supplemented with either 250 mmol/l glucose (GLU, n = 9) or 250 mmol/l mannitol (MAN, n = 10). Control rats did not undergo catheter implantation or the dialysis procedure (CON, n = 6). At the end of the study, a 1-hour peritoneal equilibration test (PET) using Dianeal 3.86% (Baxter Healthcare SA, Castlebar, Ireland) was performed in every dialyzed rat to analyze peritoneal transport. Afterward, the rats were humanely killed by bleeding, and a semi-quantitative scale was used to evaluate adhesions in the peritoneal cavity. Imprints of the visceral mesothelium and samples of the visceral peritoneum (liver) were then taken and analyzed by light microscopy. The PET results for glucose, urea, creatinine, and total protein were comparable in both experimental groups. We found that intraperitoneal adhesions were more severe in the MAN group (6 rats with adhesions graded >3) than in the GLU group (only 1 such rat). No difference in peritoneal thickness was observed between the experimental groups (MAN: 54.9 ± 17.8 µm; GLU: 51.2 ± 14.5 µm); however, in both experimental groups, the thickness was greater than in the CON group (3.9 ± 0.6 µm). The density of the peritoneal blood vessels tended to be greater in the From: Department of Pathophysiology, University Medical School of Poznan, Poznan, Poland, and 1 Division of Nephrology, University of Toronto, Toronto, Ontario, Canada. MAN group than in the GLU group (0.158 ± vessels/1000 µm 2 vs ± vessels/1000 µm 2, p = ). No visible blood vessels were evident in the CON group. The density of mesothelial cells was higher in the MAN group than in the GLU group (2456 ± 333 cells/mm 2 vs ± 322 cells/mm 2, p < 0.05), and, in both experimental groups, the cell density was higher than in the CON group (817 ± 100 cells/mm 2, p < 0.01). The nucleus:cytoplasm area ratio in mesothelial cells was comparable in the MAN and GLU groups (0.206 ± and ± 0.045), but that ratio was higher in both experimental groups than in the CON group (0.086 ± 0.010, p < 0.01). We conclude that glucose-induced changes in the peritoneum of rats exposed to chronic peritoneal dialysis depend on both osmotic and metabolic effects. Key words Glucose, osmolality, peritoneum, morphology Introduction Chronic exposure of the peritoneum to peritoneal dialysis fluids (PDFs) results in alterations in permeability that progress with time and that may be the cause of patient withdrawal from peritoneal dialysis therapy (1,2). Those functional disturbances are accompanied by morphologic changes in peritoneal membrane, including mesothelial cell damage and disorganization of the peritoneal interstitium (3,4). Long-term exposure to PDFs may also result in impairment of natural host defenses and may cause increased susceptibility to peritonitis (5,6). All of those events indicate the bioincompatibility of PDFs. Glucose is known to be an important factor in the bioincompatibility of commercially available PDFs. At least two possible mechanisms are involved in the negative impact of glucose on peritoneal function and structure: a direct metabolic effect of the high glucose concentration, and hyperosmolality of the solu-

2 16 tion as a consequence of the high glucose concentration. Moreover, the glucose degradation products (GDPs) present in heat-sterilized PDFs may intensify the disadvantageous effects of peritoneal dialysis (7,8). On the other hand, repeated infusions of any fluid into the peritoneal cavity may cause an inflammatory reaction owing to mechanical irritation of peritoneum (9,10). In previous studies, we found that isotonic, glucose-free solutions such as physiologic saline and phosphate-buffered saline cause inflammatory reactions in the peritoneal cavity of rats (11,12) and, as a consequence, alterations in the morphology of the peritoneum (13). Interestingly, when glucose was added to the above-mentioned solutions, we observed a reduction in the severity of inflammation and of morphologic changes. Our aim in the present study was to determine whether the morphologic effects of glucose depend on its metabolic action or on the hyperosmolality of the glucose dialysis solution. Materials and methods The study was performed on 25 male Wistar rats weighing between 250 g and 350 g. Peritoneal catheters were implanted into 19 rats according to a previously described method (14). The control group (CON) consisted of 6 rats that did not undergo catheter implantation and dialysis. For the first two days, all rats with implanted catheters were injected intraperitoneally with Dianeal 1.36% (Baxter Healthcare SA, Castlebar, Ireland) supplemented with antibiotics [gentamicin 5 mg/l (Polfa, Tarchomin, Poland) and cefazolin 50 mg/l (Eli Lilly, Florence, Italy)]. The fluid was allowed to absorb. On the third day, the rats with catheters were randomly divided into two groups and were infused twice daily for a further 30 days with 20 ml of one of the following solutions (also supplemented with antibiotics): Group GLU (n = 9) 0.9% NaCl (Baxter Healthcare, Warsaw, Poland) supplemented with 250 mmol/l glucose (Pilva, Krakow, Poland) Group MAN (n = 10) 0.9% NaCl (Baxter Healthcare, Warsaw) supplemented with 250 mmol/l mannitol (Baxter Healthcare, Warsaw). At the end of the study, a 1-hour peritoneal equilibration test (PET) was performed. The rats were then Styszynski et al. humanely killed by bleeding, under ether anesthesia. The peritoneal cavity was opened to estimate the presence and severity of intraperitoneal adhesions. Imprints of mesothelial cells from the visceral peritoneum of the liver surface were taken to characterize cell morphology. Liver samples were also collected for morphology analysis of the visceral peritoneum. Peritoneal equilibration test Each animal was injected with 30 ml Dianeal 3.86% (Baxter Healthcare, Castlebar). Immediately after infusion and after 30 minutes of dwell time, 2 ml samples of dialysate were obtained. After 1 hour of dwell time, the remaining dialysate was completely drained and its volume was measured. At the same time, blood samples were taken for serum analysis. Peritoneal transport was determined using the dialysate-to-plasma (D/P) ratios of urea nitrogen, creatinine, and total protein, and the final-to-instillation (D/D 0 ) ratio of glucose concentration. In the samples, the urea nitrogen and creatinine concentrations were determined by an enzymatic method (kit numbers A-371 and A-291: Analco, Warsaw, Poland). Total protein concentration was measured by the Lowry colorimetric method (15). Glucose in dialysate samples was determined by the glucose oxidase peroxidase method (Sigma, St. Louis, MO, U.S.A.). Evaluation of peritoneal adhesions Intensity of the peritoneal adhesions was evaluated according to 4-point scale: 1 point Thin adhesion (1 2 mm) between catheter and peritesticular fat pad 2 points Medium adhesion (3 5 mm) between catheter and peritesticular fat pad 3 points Thick adhesion (>5 mm) between catheter and peritesticular fat pad 4 points Adhesions between intraperitoneal organs In cases where the peritoneal cavity of a rat showed more than one adhesion, we totaled the point score for the various adhesions. Mesothelial cell imprints Imprints of the mesothelial monolayer from the visceral peritoneum were taken after the peritoneal cavity was opened. Glass slides coated with 1% agar

3 Glucose Impact on Peritoneal Morphology (Sigma, St. Louis, MO, U.S.A.) were applied to the liver surface for 30 seconds, allowing mesothelial cells to be peeled off that surface. Next, the cells were fixed using 96% ethanol. Slides were stained with hematoxylin (Quimica Clinica Aplicada, Amposta, Spain). The density of the mesothelial cells and the nucleus:cytoplasm surface ratio (an index of mesothelial hyperplasia) were determined by light microscopy (observed at 200 magnification) supported with digital image analysis equipment (Lucia 4.6, Prague, Czech Republic). Visceral peritoneum morphology Liver samples were taken and fixed with 10% formaldehyde solution in phosphate-buffered saline (PBS). Paraffin-embedded tissue sections were stained using the van Gieson method for visualization of collagen. The thickness of the peritoneal membrane and the density of the blood vessels (expressed as the number of blood vessel sections per 1000 µm 2 in that membrane) were measured using the equipment mentioned earlier. Statistical analysis Results are expressed as mean ± standard deviation. The statistical analyses were performed using the Mann Whitney test. A p value of less than 0.05 was considered significant. Results We found no significant differences between the experimental groups in peritoneal transport of measured substances. Also, no differences were observed in the prevalence of intraperitoneal adhesions between the GLU and MAN groups (78% and 60% of the rats in those groups had intraperitoneal adhesions). However, the severity of the adhesions was less in the GLU group than in the MAN group. In the GLU group, only 1 rat had adhesions evaluated above 3 points. In the MAN group, 6 rats had adhesions evaluated above 3 points. We found no adhesions in the CON group. Animals from the GLU group also had a lower density of mesothelial cells than rats in the MAN group [2090 ± 322 cells/mm 2 vs ± 333 cells/mm 2, p < 0.05; Figure 1(C)]. Both experimental groups had a higher density of mesothelial cells than the CON group (817 ± 100 cells/mm 2, p < 0.01). No significant difference was observed in the nucleus:cytoplasm area ratio between the GLU group and the MAN group 17 (0.176 ± vs ± 0.039); but, in both experimental groups, that ratio was higher than in the CON group [0.086 ± 0.010, p < 0.01; Figure 1(D)]. Peritoneal thickness was comparable in the GLU and MAN groups (51.2 ± 14.5 µm vs ± 17.8 µm) and, in both experimental groups, exceeded the thickness in the CON group [3.9 ± 0.6 µm, p < 0.01; Figure 1(A)]. We noted a tendency for the density of the blood vessels in the visceral peritoneum to be lower in the GLU group than in the MAN group [0.085 ± vessels/1000 µm 2 vs ± vessels/ 1000 µm 2, p = ; Figure 1(B)]. No visible vessels were observed in the peritoneal membrane of rats from the CON group. Discussion In previous experiments, we found that glucose suppresses inflammatory reactions (measured as cell count, protein, nitric oxide, monocyte chemoattractant protein-1 concentration) and morphologic alterations in the peritoneal cavity (mesothelial hyperplasia, fibrosis, and angiogenesis) induced by chronic intraperitoneal infusions of iso-osmotic saline (12,13). In the present experiment, we tried to determine whether the unexpected impact of glucose on peritoneal morphology is related solely to its osmotic action or to its metabolic activity. Certain effects of chronic exposure to glucose and mannitol seemed to be comparable in both experimental groups (peritoneal transport of urea, creatinine, protein, and glucose; peritoneal thickness; prevalence of adhesions). But some of the observed changes were more intensive in animals that were exposed to mannitol (as compared with animals that were administered glucose in the same concentration). Those changes include mesothelial cell density (expressed as a higher intensity of mesothelial hyperplasia in the MAN group) and peritoneal vessel density (expressed as increased angiogenesis in the MAN group). Moreover, among rats in which intraperitoneal adhesions were present, the adhesions were more severe if mannitol was used as the osmotic solute in the dialysis fluid. Those findings lead us to conclude that glucoseinduced changes in the morphology of the peritoneal cavity (mesothelial hyperplasia, angiogenesis, formation of peritoneal adhesions) are less intense than those caused solely by osmotic action (mannitol). This situation suggests that unidentified meta-

4 18 Styszynski et al. FIGURE 1 Morphologic changes in peritoneal cavity of rats exposed to 0.9% NaCl supplemented with glucose 250 mmol/l (GLU) or mannitol 250 mmol/l (MAN) and in non dialyzed rats (CON). * p < 0.05 as compared with CON; ** p < 0.01 as compared with CON; + p < 0.05 as compared with MAN; # p = as compared with MAN. bolic effects of glucose may modify certain morphologic changes in the peritoneum under conditions of chronic dialysis. Reduced proliferation of mesothelial cells in a high glucose environment (as compared with the same concentration of mannitol) was previously demonstrated by in vitro experiments (16,17). Gotloib et al. (18) also observed a lower density of mesothelial cells in mice intraperitoneally injected with high-glucose PDF as compared with mice injected with another osmotic solution (polyglucose). Those researchers explained their finding as acceleration of the lifecycle of mesothelial cells following depletion of the growth capabilities of the monolayer exposed to a high glucose solution (19). Thus, the lower density of mesothelial cells in rats injected with high glucose as compared with rats injected with mannitol may reflect an ineffective repair mechanism after injury induced by dialysis. On the other hand, the lower intensity of morphologic change in the peritoneal cavity may be a sign of a diminished inflammatory reaction, which may result from depletion of peritoneal macrophages caused by toxic glucose activity against those cells (5). To verify that hypothesis, the antibacterial activity of peritoneal macrophages needs to be examined. Conclusion Glucose-induced changes in the peritoneum of rats exposed to chronic peritoneal dialysis depend on both the osmotic and the metabolic effects of glucose. Those changes are less intense than the changes caused by a PDF of the same osmolality containing mannitol. This unexpected finding may arise from a glucose-depen-

5 Glucose Impact on Peritoneal Morphology dent suppressive influence on the regenerative and antibacterial potentials of the peritoneum. References 1 Davies SJ, Bryan J, Phillips L, Russell GI. Longitudinal changes in peritoneal kinetics: the effect of peritoneal dialysis and peritonitis. Nephrol Dial Transplant 1996; 11: Kawaguchi Y, Hasegawa T, Nakayama M, Kubo H, Shigematsu T. Issues affecting the longevity of the continuous peritoneal dialysis therapy. Kidney Int 1997; 52(Suppl 62):S Di Paolo N, Sacchi G, De Mia M, et al. Morphology of the peritoneal membrane during continuous ambulatory peritoneal dialysis. Nephron 1986; 44: Dobbie JW, Anderson JD, Hind C. Long-term effects of peritoneal dialysis on peritoneal morphology. Perit Dial Int 1994; 14(Suppl 3):S Liberek T, Topley N, Jörres A, Coles GA, Gahl GM, Williams JD. Peritoneal dialysis fluid inhibition of phagocyte function: effects of osmolality and glucose concentration. J Am Soc Nephrol 1993; 3: Cendoroglo M, Sundaram, Groves C, Ucci AA, Jaber BL, Pereira BJ. Necrosis and apoptosis of polymorphonuclear cells exposed to peritoneal dialysis fluids in vitro. Kidney Int 1997; 52: Nilsson Thorell CB, Muscalu N, Andrén AH, Kjellstrand PT, Wieslander AP. Heat sterilization of fluids for peritoneal dialysis gives rise to aldehydes. Perit Dial Int 1993; 13: Wieczorowska Tobis K, Polubinska A, Schaub TP, et al. Influence of neutral-ph dialysis solutions on the peritoneal membrane: a long-term investigation in rats. Perit Dial Int 2001; 21(Suppl 3):S Dobbie JW. Pathogenesis of peritoneal fibrosing syndromes (sclerosing peritonitis) in peritoneal dialysis. Perit Dial Int 1992; 12: Wang T, Heimbürger O, Qureshi AR, Waniewski J, Bergström J, Lindholm B. Physiological saline is not 19 a biocompatible peritoneal dialysis solution. Int J Artif Organs 1999; 22: Wieczorowska Tobis K, Styszynski A, Polubinska A, Radkowski M, Breborowicz A, Oreopoulos DG. Hypertonicity of dialysis fluid suppresses intraperitoneal inflammation. Adv Perit Dial 2000; 16: Styszynski A, Podkowka R, Wieczorowska-Tobis K, et al. Glucose suppresses peritoneal reactions and mesothelial hyperplasia caused by intraperitoneal saline infusion. Adv Perit Dial 2002; 18: Styszynski A, Kwiatkowska B, Podkowka R, Wieczorowska Tobis K, Breborowicz A, Oreopoulos DG. Impact of morphological alterations caused by chronic dialysis in rats on peritoneal transport [Abstract]. Perit Dial Int 2002; 22: Moore HL. Chronic catheter model for exchanges in the ambulatory rat [Abstract]. Perit Dial Int 1992; 12(Suppl 2):S Lowry OH, Rosenbrough NJ, Rarr AL, Randall RJ. Protein measurement with the folin phenol reagent. J Biol Chem 1951; 193: Breborowicz A, Rodela H, Oreopoulos DG. Toxicity of osmotic solutes on human mesothelial cells in vitro. Kidney Int 1992; 41: Ito T, Yorioka N, Yamamoto M, Kataoka K, Yamakido M. Effect of glucose on intercellular junctions of cultured human peritoneal mesothelial cells. J Am Soc Nephrol 2000; 11: Gotloib L, Shostak A, Wajsbrot V. Effects of osmotic agents upon the in vivo exposed mesothelial monolayer. Perit Dial Int 2000; 20(Suppl 2):S Shostak A, Wajsbrot V, Gotloib L. High glucose accelerates the life cycle of the in vivo exposed mesothelium. Kidney Int 2000; 58: Corresponding author: Arkadiusz Styszynski, MD, Department of Pathophysiology, University Medical School of Poznan, ul. Swiecickiego 6, Poznan Poland.

6 Advances in Peritoneal Dialysis, Vol. 19, 2003 Katarzyna Wieczorowska Tobis, 1 Renata Brelinska, 2 Andrzej Breborowicz, 1 Dimitrios G. Oreopoulos 3 Morphologic Changes in the Liver of Dialyzed Rats Preliminary Observations Peritoneal dialysis affects both the quantity and quality of connective tissue in the visceral peritoneum. In the present study, we report the alterations observed in the morphology of the superficial liver lobuli of dialyzed rats. The studies were performed in male Wistar rats weighing g. Rats were exposed intraperitoneally to 0.9% NaCl, phosphate-buffered saline (PBS), or commercial dialysis solutions containing 3.86% glucose (Dianeal 3.86%: Baxter Healthcare SA, Castlebar, Ireland; CAPD3 and CAPD3 Balance: Fresenius Medical Care, Bad Homburg, Germany) twice daily for 4 6 weeks. At the end of the study, samples of the liver were taken and stained for light microscopy (hematoxylin and eosin, van Gieson). The results obtained in dialyzed rats were compared with those from the control, non dialyzed animals. In control animals, the surface of the hepatic parenchyma was smooth. In all dialyzed rats, irrespective of the solution used, folding of the surface of the liver parenchyma was found owing to penetration of connective tissue elements between the hepatocytes. In effect, folds of hepatocytes within the liver capsule became detached and isolated from the remaining cells of the lobules. The distinctive feature of that pathologic change was that its severity increased with the thickness of the peritoneum. The change was seen in rats exposed to any of the experimental solutions, and therefore appeared to be attributable to a nonspecific reaction caused by exposure of the peritoneum to dialysis fluid and not to specific components of the fluid. The observed alterations in morphology seem to suggest disturbed function within the affected lobules. Further studies are necessary to confirm this hypothesis. From: 1 Department of Pathophysiology and 2 Department of Histology and Embryology, University of Medical Sciences, Poznan, Poland, and 3 Division of Nephrology, University of Toronto, Toronto, Ontario, Canada. Key words Connective tissue, morphology, liver Introduction Morphologic screening of the peritoneum of dialyzed patients has permitted characterization of the typical findings in parietal peritoneum attributable to the dialysis procedure (1 3). Similarly, in animals chronically exposed to dialysis fluids, peritoneal thickening and increased vascularity of the parietal peritoneum are relatively well characterized (4 6). However, data about the impact of dialysis on the visceral peritoneum are sparse. Hypervascularity, an increased amount of extracellular matrix, and a larger number of milky spots were found in the omenta of animals exposed to standard high glucose dialysis solution (6). Previously, our interest focused on the visceral peritoneum of the liver in chronically dialyzed rats (7 9). We showed that the expansion of submesothelial tissue was attributable to an increased quantity of collagen fibers and their disorganization and disintegration; to an increased quantity of connective-tissue cells and their activation; and to increased vascularity (10). Changes in the visceral peritoneum of the liver were anatomically asymmetric, appearing mainly on the anterior and inferior surfaces of the liver, which have prolonged contact with the dialysis fluid (10). We speculated that alterations in all components of the connective tissue of the visceral peritoneum of the liver owing to dialysis might interfere with the function of the liver. We could find no literature on the possibility of liver fibrosis caused by dialysis. Our aim in the present study was therefore to report abnormalities observed in the morphology of the anterior and inferior surfaces of the liver in dialyzed rats. Materials and methods Samples of visceral peritoneum of the liver from rats exposed to various dialysis solutions were retrospec-

7 Wieczorowska Tobis et al. 21 tively screened for liver alterations in connection with peritoneal fibrosis. Dialysis procedure The experiment was performed on male Wistar rats weighing g. All animals were maintained in standard conditions during the study period (temperature: 18 ± 1 C; 12/12 light/dark cycles), were fed a standard rat diet, and had free access to water. After catheter implantation, the experimental animals were intraperitoneally exposed twice daily for 4 6 weeks to 0.9% NaCl, to phosphate-buffered saline (PBS), or to commercial dialysis solutions containing 3.86% glucose (Dianeal 3.86%: Baxter Healthcare SA, Castlebar, Ireland; CAPD3 and CAPD3 Balance: Fresenius Medical Care, Bad Homburg, Germany). All fluids were supplemented with antibiotics [gentamicin 5 mg/l (Polfa, Tarchomin, Poland) and cefazolin 50 mg/l (Eli Lilly, Florence, Italy)]. Control animals were exposed to neither catheter implantation nor dialysis procedure. Specimen collection and staining At the end of the experiment, all rats were humanely killed by bleeding. The peritoneal cavity was opened, and samples of the visceral peritoneum of the liver were taken with a minimum of handling or trauma. The samples were fixed in 1% phosphate-buffered formaldehyde, and sections (4 5 µm in thickness) were stained with hematoxylin and eosin and van Gieson (staining for collagen, which appears reddish). Results In control animals, the surface of the hepatic parenchyma was smooth. It was covered by a thin layer of connective tissue and a monolayer of mesothelial cells, forming the liver capsule [Figure 1(A)]. In dialyzed rats, folding of the surface of the liver parenchyma was observed. The folding was attributable to penetration of connective tissue elements between the hepatocytes. In effect, tangled knots of liver cells were found within the connective tissue of the liver capsule [Figure 1(B)]. Those cells were usually in close proximity to the vessels [Figure 2(A,B)]; however, the possible connections between neovascularization of the peritoneum and the presenting phenomenon are not clear. We observed groups of hepatocytes only partially surrounded by fibrous tissue [Figure 2(C)]. But, on FIGURE 1 Sections of the visceral peritoneum of the liver (L). (A) Control animal: The smooth surface of the hepatic parenchyma is covered by a thin capsule (arrowheads point to the nuclei of mesothelial cells), 200. (B) Dialyzed animal: Tangled knots of liver cells are seen within the connective tissue of the liver capsule (black arrowheads), 200. The white double-headed arrow marks the expanded peritoneum. the other hand, we also found cells encircled by a separate layer of fibrous tissue, which formed a sort of connective tissue capsule for the hepatocytes that were detached and isolated from the remaining cells of the lobulus [Figure 2(D)]. Isolated cells showed features of glycogen degeneration. It is not clear whether the presenting alterations can be interpreted as various stages in the expansion of connective tissue or simply as folding of the various sections. Composition of the dialysis fluid appeared to have no impact on the presenting phenomenon. Similar alterations were found in rats exposed to each of the studied solutions. However, a distinctive feature of

8 22 Morphology Changes in the Liver FIGURE 2 Sections of the visceral peritoneum from the liver (L) of dialyzed animals. The white double-headed arrows mark the expanded peritoneum. (A, B) Hepatocytes are detached and isolated from the remaining cells of the lobulus (black arrowheads) in near proximity of the vessels (V). (A): 400 ; (B): 200. (C) Folding of the surface of the liver parenchyma. The group of hepatocytes is only partly surrounded by fibrous tissue (black arrowhead), 200. (D) Liver cells (black arrowhead) encircled by a separate layer of fibrous tissue, 600. the pathologic changes was that the thicker the submesothelial tissue was, the more likely were alterations to be found. The alterations were always present on the anterior and inferior surfaces of the liver where thickening was the most severe. Discussion Peritoneal dialysis is well known to affect both the quantity and the quality of connective tissue in the parietal peritoneum of chronically dialyzed patients and animals (1 6). Owing to severe fibrosis of the visceral peritoneum of the liver, it seems possible that exposure to dialysis fluids affects the biologic properties of the hepatocytes in the superficial liver lobuli. Interestingly, subcapsular steatonecrosis in the liver has previously been reported in dialyzed patients with diabetes as an effect of intraperitoneal treatment with insulin (11,12). The presenting alterations do not seem to be caused by the high glucose concentration and hyperosmolality of the dialysis fluid, because they were also present in rats dialyzed with isotonic glucose-free solutions (0.9% NaCl, PBS). Additionally, the alterations are not explained by low ph, because they were also present in animals exposed to dialysis fluids with a normal ph (PBS, CAPD3-Balance). The alterations might be speculated to be caused by nonspecific irritation resulting from the exposure of the peritoneum to dialysis fluid and not to specific components of the fluid.

9 Wieczorowska Tobis et al. Conclusion In this paper, we report for the first time the folding of the surface of the liver parenchyma followed by sequestration of superficial liver cells. Because the functions of cells within the lobules are differentiated, the observed alterations in morphology seem to suggest disturbed function within the affected lobules. Further studies are necessary to confirm this hypothesis and to discover its consequences. References 1 Williams JD, Craig KJ, Topley N, et al. Morphologic changes in the peritoneal membrane of patients with renal disease. J Am Soc Nephrol 2002; 13: Dobbie JW, Anderson JD, Hind C. Long-term effects of peritoneal dialysis on peritoneal morphology. Perit Dial Int 1994; 14(Suppl 3):S Pollock CA, Ibels LS, Eckstein RP, et al. Peritoneal morphology on maintenance dialysis. Am J Nephrol 1989; 9: Duman S, Sen S, Gunal AI, Asci G, Akcicek F, Basci A. How can we standardize peritoneal thickness measurements in experimental studies in rats? Perit Dial Int 2001; 21(Suppl 3):S Margetts PJ, Kolb M, Yu L, Hoff CM, Gauldie J. A chronic inflammatory infusion model of peritoneal dialysis in rats. Perit Dial Int 2001; 21(Suppl 3): S Hekking LH, Zareie M, Driesprong BA, et al. Better preservation of peritoneal morphologic features and 23 defense in rats after long-term exposure to a bicarbonate/lactate buffered solution. J Am Soc Nephrol 2001; 12: Wieczorowska Tobis K, Brelinska R, Breborowicz A, Oreopoulos DG. Morphologic aspects of chronic peritoneal dialysis in a rat model. Perit Dial Int 2001; 21(Suppl 3):S Wieczorowska Tobis K, Brelinska R, Polubinska A, Breborowicz A. The ultrastructure of the peritoneal membrane in chronically dialysed rats. Folia Histochem Cytobiol 2001; 39: Korybalska K, Wieczorowska Tobis K, Polubinska A, et al. L-2-Oxothiazolidine-4-carboxylate: an agent that modulates lipopolysaccharide-induced peritonitis in rats. Perit Dial Int 2002; 22: Wieczorowska Tobis K, Dobbie JW, Anderson JH, et al. Peritoneal morphology in chronically dialyzed rats [Abstract]. Perit Dial Int 1999; 19(Suppl 1):S6. 11 Wanless IR, Bargman JM, Oreopoulos DG, Vas SI. Subcapsular steatonecrosis in response to peritoneal insulin delivery: a clue to the pathogenesis of steatonecrosis in obesity. Mod Pathol 1989; 2: Kallio T, Nevalainen PI, Lahtela JT, Mustonen J, Pasternack A. Hepatic subcapsular steatosis in diabetic CAPD patients treated with intraperitoneal insulin. Acta Radiol 2001; 42: Corresponding author: Katarzyna Wieczorowska Tobis, MD, Pathophysiology Department, University Medical School, ul. Swiecickiego 6, Poznan Poland.

10 Advances in Peritoneal Dialysis, Vol. 19, 2003 Sema Akman, 1 Roos van Westrhenen, 1 Dirk R. De Waart, 2 Johan K. Hiralall, 2 Machteld M. Zweers, 1 Raymond T. Krediet 1 The Effect of Dwell Time on Dialysate Cancer Antigen 125 Appearance Rates in Patients on Continuous Ambulatory Peritoneal Dialysis The dialysate concentration of cancer antigen 125 (CA125) can be considered a reflection of mesothelial cell mass or turnover in stable continuous ambulatory peritoneal dialysis (CAPD) patients. The effect of dwell times exceeding 4 hours on CA125 appearance rate (CA125AR) is not known. Therefore, our objective in the present study was to analyze the effect of dwell time on CA125AR in stable CAPD patients. In 43 stable CAPD patients, we analyzed standard peritoneal permeability analyses (SPAs) performed with a 3.86% glucose dialysate, and night-dwell effluents from the night dwell prior to the SPA. Dialysate CA125 concentration was measured by radioimmunoassay (RIA II: Fujirebio Diagnostics, Malvern, PA, U.S.A.). Night-dwell CA125 correlated with the duration of the dwell (r = 0.32, p = 0.04) and with the CA125 concentration in the 4-hour dwell (r = 0.83, p < 0.001). The mean CA125AR in the SPA effluent was 97.8 ± 46.3 U/min; in the overnight effluent, it was ± 73.7 U/min (nonsignificant). A good correlation was present between the CA125AR in the 4-hour dwells and in the overnight dwells (r = 0.82, p < 0.001). We conclude that using night dwells to regularly assess dialysate CA125 for instance, at every outpatient visit is possible in CAPD patients, provided that appearance rate is calculated. Key words Cancer antigen 125, CAPD, dwell time From: 1 Division of Nephrology, Department of Medicine, and 2 Department of Clinical Chemistry, Academic Medical Center and University of Amsterdam, Amsterdam, Netherlands. Introduction The dialysate concentration of cancer antigen 125 (CA125), a high molecular weight (220 kda) glycoprotein, can be considered a reflection of mesothelial cell mass or turnover in stable continuous ambulatory peritoneal dialysis (CAPD) patients [reviewed in (1)]. Originally, CA125 was measured in overnight effluent because such measurements yielded the highest values (2). Later, studies were performed in peritoneal effluent obtained after a standardized 4-hour dwell because dialysate CA125 showed a linear increase during the 4 hours of the dwell regardless of the dialysate glucose concentration used (3). Standardization of dialysate CA125 measurements might therefore be achieved by using a fixed dwell time in every patient or by calculating the CA125 appearance rate (CA125AR) that is, the amount of CA125 present in the total drained effluent divided by the duration of the dwell in minutes. The latter approach would make it possible to use overnight dwells of varying duration and still obtain accurate results that can be used for follow-up. However, the effect of dwell times exceeding 4 hours on the CA125AR was not known. Therefore, our objective in the present study was to analyze the effect of dwell time on CA125AR in stable CAPD patients. Patients and methods In 43 stable CAPD patients, we analyzed standard peritoneal permeability analyses [SPAs (4)] performed with 3.86% glucose dialysate, and night-dwell effluents from the night dwell prior to the SPA. The mean age of the patients was 55 years (range: years) and the mean duration of CAPD was 29 months (range: months). The incidence of peritonitis

11 Akman et al. was 1 episode per 38 patient months. None of the patients had peritonitis at the time of the SPA or during the preceding 4 weeks. The dwell time of each SPA was, by definition, 240 minutes. The dwell time of the night dwells ranged from 220 minutes to 910 minutes (mean: 587 minutes). Dialysate samples were obtained prior to the SPA and after completion of the SPA. The samples were centrifuged and stored at 20 C until analysis. Dialysate CA125 concentrations were measured by radioimmunoassay using two monoclonal antibodies namely, OC125 and M11 (RIA II: Fujirebio Diagnostics, Malvern, PA, U.S.A.). The dialysate appearance rates of CA125 were calculated by multiplying the dialysate concentration by the drained volume and dividing the result by the duration of the dwell in minutes. The CA125ARs are expressed as units per minute. All data are presented as mean ± standard deviation. Comparisons between the 4-hour dwells and the overnight dwells used the paired samples t-test. Correlations were analyzed by linear regression analysis. In addition, the CA125ARs were compared using the Bland Altman method (5). Results The mean CA125 concentration in SPA effluent was 12.1 ± 5.7 U/mL; in the overnight effluent, it was 31.4 ± 20.9 U/mL (p < 0.001). Figure 1 shows individual data. The average CA125 concentration showed a linear increase with the duration of the dwell, as illustrated in Figure 2. Night-dwell CA125 correlated with the duration of the dwell (r = 0.32, p = 0.04) and with the CA125 concentration in the 4-hour dwell (r = 0.83, p < 0.001). The mean CA125AR in the SPA effluent was 97.8 ± 46.3 U/min; in the overnight effluent, it was ± 73.7 U/min (nonsignificant). Figure 3 shows the individual data. Although the mean values were not significantly different, the coefficient of inter-individual variation (standard deviation / mean 100) was 47% for the 4-hour dwell and 68% for the overnight dwell. Good correlation was present between the CA125AR in the 4-hour dwells and in the overnight dwells (r = 0.82, p < 0.001), as shown in Figure 4. The Bland Altman analysis (Figure 5) showed that the mean difference between the 4-hour dwells and the overnight dwells was only U/min, but that the CA125ARs calculated for the 4-hour dwells 25 FIGURE 1 A comparison between the concentrations of cancer antigen 125 (CA125) obtained in dialysate after a 4-hour dwell and after the preceding overnight dwell. The lines connect the data from individual patients. FIGURE 2 Mean concentrations of cancer antigen 125 (CA125) in dialysate after 4, 8, 10, and 12 hours. tended to be higher than those obtained for the overnight dwell for low values, and tended to be lower for higher values. The standard deviation of the difference between the two methods was 44.8 U/min, resulting in limits of agreement between 100 U/min and 80 U/min. Discussion Mesothelial cells are a biological barrier against potentially bioincompatible dialysis fluid and are also known to be involved in local host defense. In addi-

12 26 Dwell Time and CA125 Appearance Rates in CAPD FIGURE 3 A comparison of the appearance rates of cancer antigen 125 (CA125) in dialysate obtained after a 4-hour dwell and after the preceding overnight dwell. The lines connect the data from individual patients. FIGURE 5 Bland Altman analysis of appearance rates (ARs) of cancer antigen 125 (CA125) obtained after a 4-hour dwell and after the preceding overnight dwell. The solid horizontal line shows the mean value of the difference. The broken lines represent the 95% confidence limits. The overnight value tends to underestimate the 4-hour value in the low range and to overestimate it in the high range. found no significant differences between 4-hour and long dwells when we calculated CA125AR. Furthermore, a strong correlation was evident between dwells of both lengths. FIGURE 4 Correlation of the appearance rates of cancer antigen 125 (CA125) in dialysate from the preceding night dwell (ND). Each dot represents an individual patient. The linear regression is CA125 (4 h) = 0.51 CA125 ND (r = 0.82, p < 0.001). tion, they produce substances that prevent the formation of adhesions. The glycoprotein CA125 is a high molecular weight (220 kda) protein produced by mesothelial cells. Its effluent concentration in stable PD patients is likely to reflect mesothelial cell mass or turnover (2,6). However, the method used to evaluate dialysate CA125 levels has not yet been standardized. Dialysate CA125 increases linearly during the first 4 hours of a dialysis dwell (3,7). The present study showed that the linear increase was sustained during dwells of up to 12 hours. That finding suggests that long dwells can also be used in analyzing CA125, because, provided that the duration of the dwell is known, the CA125AR can be calculated. Indeed, we Conclusion We conclude that night dwells can be used for the regular assessment of dialysate CA125 in CAPD patients for instance, at every outpatient visit provided that appearance rate is calculated. The method detects downward trends that suggest loss of mesothelial cell mass. Although risk concentrations of CA125 have not yet been defined, it is likely that a downward trend indicates the presence of morphologic change in the peritoneal membrane. References 1 Krediet RT. Dialysate cancer antigen 125 concentration as marker of peritoneal membrane status in patients treated with chronic peritoneal dialysis. Perit Dial Int 2001; 21: Koomen GC, Betjes MG, Zemel D, Krediet RT, Hoek FJ. Cancer antigen 125 is locally produced in the peritoneal cavity during continuous ambulatory peritoneal dialysis. Perit Dial Int 1994; 14: Ho-dac-Pannekeet MM, Hiralall JK, Struijk DG, Krediet RT. Longitudinal follow-up of CA125 in

13 Akman et al. peritoneal effluent. Kidney Int 1997; 51: Pannekeet MM, Imholz AL, Struijk DG, et al. The standard peritoneal permeability analysis: a tool for the assessment of peritoneal permeability characteristics in CAPD patients. Kidney Int 1995; 48: Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986; 1: Visser CE, Brouwer Steenbergen JJ, Betjes MG, Koomen GC, Beelen RH, Krediet RT. Cancer antigen 125: a bulk marker for the mesothelial mass in stable peritoneal dialysis patients. Nephrol Dial Transplant 1995; 10: Jimenez C, Diaz C, Selgas R, et al. Peritoneal kinetics of cancer antigen 125 in peritoneal dialysis patients: the relationship with peritoneal outcome. Adv Perit Dial 1999; 15: Corresponding author: Raymond T. Krediet, MD, Academic Medical Center, Renal Unit, F4-215, Meibergdreef 9, Amsterdam 1105 AZ, Netherlands.