Peritoneal Transport Characteristics with Glucose Polymer– Based Dialysis Fluid in Children

Home , Icodextrin

J Am Soc Nephrol 15: 2940–2947, 2004 Peritoneal Transport Characteristics with Glucose Polymer– Based Dialysis Fluid in Children

ESTHER RUSTHOVEN,* RAYMOND T. KREDIET,† HANS L. WILLEMS,‡ LEO A.H. MONNENS,§ and CORNELIS H. SCHRO¨ DER* *Department of Pediatric Nephrology, University Medical Center Utrecht, Utrecht, The Netherlands; †Department of Nephrology, Academic Medical Center, Amsterdam, The Netherlands; ‡Department of Clinical Chemistry, University Medical Center St. Radboud, Nijmegen, The Netherlands; and §Department of Pediatric Nephrology, University Medical Center St. Radboud, Nijmegen, The Netherlands

Abstract. Scarce data are available on the use of glucose tions, except for TCUF, which was lower for icodextrin (0.9 polymer–based dialysate in children. The effects of glucose ml/min per 1.73 m2) as compared with 3.86% glucose (4 polymer–based dialysate on peritoneal fluid kinetics and solute ml/min per 1.73 m2). Transport parameters were similar in transport were studied in pediatric patients who were on children and adults for glucose, but with icodextrin, TCUF and chronic peritoneal dialysis, and a comparison was made with marker clearance were significantly lower in children. AQP- previously published results in adult patients. In nine children, mediated water flow was 83 versus 50% with glucose (child two peritoneal equilibration tests were performed using 3.86% versus adult; P Ͻ 0.01) and 18 versus 7% with icodextrin (P Ͻ glucose and 7.5% icodextrin as a test solution. Dextran 70 was 0.01). Data indicate that transport parameters in children using added as a volume marker to calculate fluid kinetics. Serum icodextrin are similar to glucose except for TCUF. Differences and dialysate samples were taken for determination of urea, are explained by the absence of crystalloid osmosis and that creatinine, and sodium. After calculation of the initial trans- TCUF was determined after a 4-h dwell. Comparison of transcapillary ultrafiltration (TCUF) rate, it was possible to calcu- port parameters and peritoneal membrane characteristics be- late the contribution of aquaporin-mediated (AQP-mediated) tween children and adults reveal that there seem to be differ- water transport to ultrafiltration for icodextrin and 3.86% glu- ences in the amount and functionality of AQP. However, there cose and the part of LpS (the product of the peritoneal surface are no differences in clinical efficacy of this transport pathway area and the hydraulic permeability) caused by AQP. In chil- because the absolute flow through the AQP is identical in both dren, the transport parameters were similar for the two solu- groups using 3.86% glucose.

Icodextrin contains glucose polymers as osmotic agent instead icodextrin-based dialysis solution with a 3.86% glucose solu- of glucose in the conventional peritoneal dialysis (PD) solution with regard to peritoneal fluid kinetics and solute transport tions. The effectiveness of icodextrin as a colloid osmotic in a pediatric PD population. In addition, the results obtained in agent has been very well established in adult patients (1–4). children were compared with results obtained in adult patients, The solution is especially indicated in situations in which a which were previously published (5). A brief summary of the high exposure to glucose should be avoided, such as in patients study in adult patients is given in the Materials and Methods with exchanges with a long dwell time and in patients with section. ultrafiltration failure (5). Until now, very little has been published about the use of glucose polymer–based dialysate in children. It was demon- Materials and Methods strated that 7.5% icodextrin is capable of inducing sustained Pediatric Study net ultrafiltration during long-term dwell in children and that The patient group consisted of four girls and five boys, with a median age of 4.9 yr (range 1.6 to 10.9). The mean duration of nightly the metabolism of icodextrin is similar compared with adults intermittent PD treatment was 26.2 mo (range 5.6 to 122.3). In each (6). The aim of the present study was to compare a 7.5% patient, two peritoneal equilibration tests were performed, using a different test solution for each peritoneal equilibration test. The solutions used were a 7.5% icodextrin solution (Extraneal; Baxter B.V., Received December 24, 2002. Accepted August 17, 2004. Utrecht, The Netherlands) and a 3.86% glucose solution (Dianeal; Correspondence to Dr. Esther Rusthoven, Department of Pediatric Nephrol- Baxter B.V.). All peritoneal equilibration tests were performed as ogy, UMCU, Wilhelmina Children’s Hospital, KE 04.133.1, PO Box 85090, described previously by Reddingius et al. (7). An intraperitoneal 3508 AB Utrecht, The Netherlands. Phone: ϩ31-30-2504001; Fax: ϩ31-30- volume of 1200 ml/m2 body surface area (BSA) was used in all tests. 2505349; E-mail: [email protected]. Dextran 70 (Macrodex NPBI; Emmercompascuum, The Netherlands) 1046-6673/1511-2940 was added to the dialysate as a volume marker to calculate fluid Journal of the American Society of Nephrology kinetics. A serum sample was taken at the start of the study. Dialysis Copyright © 2004 by the American Society of Nephrology fluid was sampled before inflow; after 5, 30, 60, 120, and 180 min, DOI: 10.1097/01.ASN.0000143742.48705.7B and at the end of the test at 240 min. These samples were used for J Am Soc Nephrol 15: 2940–2947, 2004 Glucose Polymer–Based Dialysis Fluid in Children 2941 measurement of dextran, glucose, creatinine, urea, and sodium. All Statistical Analyses peritoneal equilibration tests were performed at least 2 mo after any Results are given as mean and median values, SD, and ranges. For peritonitis episode. None of the patients had ultrafiltration failure. comparison of the results of the two solutions within the pediatric group, a paired t test was performed. Differences between children and adults were tested with the Mann-Whitney nonparametric rank Calculations test. Correlations were tested using the Spearman rank correlation For the calculations, the principles of Nolph et al. (8) were applied, analysis. adapted by Krediet et al. (9). Transport parameters were calculated according to previously described formulas (7). In brief, transcapillary ultrafiltration (TCUF) was calculated from the dilution of the volume Study in Adult Patients marker by subtracting the initial theoretical intraperitoneal volume Ho-Dac-Pannekeet et al. (5) previously published a study about (IPV) from the theoretical IPV. The theoretical IPV is the IPV in the peritoneal transport characteristics with icodextrin performed in absence of marker clearance and sampling, in which marker clearance adults. Results obtained in this study were used (with permission of equals the disappearance of fluid from the peritoneal cavity. The Ho-Dac-Pannekeet et al.) to make a comparison with the results of our initial TCUF for the glucose solution and the icodextrin solution, study, which was performed in children. The patient group of Ho- Dac-Pannekeet et al. consisted of 10 stable patients, with a median meaning the TCUF during the first minute of a dwell (TCUF0–1min), were calculated according to the Lineweaver-Burk plot for the glu- age of 48 yr (range 23 to 64). The mean duration of continuous cose-based solution and by linear regression for the icodextrin-based ambulatory PD treatment was 28 mo (range 3 to 92). In each patient, one (9). The TCUF rate (TCUFR) was obtained by dividing TCUF by three peritoneal equilibration tests were performed, using a different the dwell time. test solution for each peritoneal equilibration test. The three test The change in IPV (⌬-IPV) was obtained by calculating the dilu- solutions consisted of 1.36% glucose, 3.86% glucose, and 7.5% ico- tion of the volume marker after correction for incomplete recovery. dextrin. The peritoneal equilibration test was standardized in the same The net ultrafiltration rate was obtained by dividing ⌬-IPV by the way as in the pediatric study. Dialysate samples were taken at 10, 20, dwell time. Marker clearance was defined as the difference between 30, 60, 120, 180, and 240 min. Blood samples were drawn at the the amount of dextran instilled and the total amount recovered, di- beginning and at the end of the period. In the glucose dwells, dextran vided by the product of dwell time and the mean dextran concentra- 70 was added as a volume marker, whereas in the icodextrin dwell, tion. Marker clearance rate was calculated by dividing marker clear- dextrin itself was used for that purpose. Calculations of transport ance by the dwell time. It was assumed that marker clearance is a parameters were made on the basis of the same principles as those linear process. The mass transfer area coefficient (MTAC) is the used in the pediatric study. Calculations of the fractional transcellular maximal theoretical diffusive clearance of a solute at time 0, before UFC and the fractional osmotic force across AQP were not part of the transport has actually started. The MTAC of urea and creatinine were adult study. calculated according to the Waniewski model (10), in which a correction for plasma water concentrations was used (5). The dialysate/ plasma ratio of sodium was used to analyze the sieving of sodium Results during the first hour of the dwell for the 3.86% glucose and the 7.5% The medians and ranges of fluid and solute transport param- icodextrin peritoneal equilibration test. eters of children and adults Glucose induces ultrafiltration by increasing the crystalloid os- obtained with 3.86% glucose and 7.5% icodextrin are given motic pressure in the peritoneal cavity, which induces fluid transport in Table 1. All data are expressed per 1.73 m2 BSA. across the small interendothelial pores and also through the ultrasmall transcellular pores (aquaporins [AQP]). The effect of glucose on the large pores can be neglected because of their very small number and Fluid Transport large pore size. The colloid osmotic pressure induced by icodextrin The TCUFR with icodextrin was significantly lower com- almost exclusively exerts its effect across the small pores. This sug- pared with the TCUFR obtained with glucose 3.86% in our gests that the ultrafiltration coefficient (UFC) of the transcellular study group (P Ͻ 0.001). The marker clearance rate and net pores (UFCaqp) can be calculated from the difference between the ultrafiltration rate were similar for the two solutions. Fluid total UFC (UFCtot) of the peritoneum, as calculated with 3.86% profiles for 3.86% glucose and 7.5% icodextrin are given in glucose, and the UFC of the small pores (UFCsp), as calculated with Figure 1. ϭ Ϫ icodextrin: UFCaqp UFCtot UFCsp. Transport parameters for fluid transport using 3.86% glu- The UFC was calculated as described previously by Ho-Dac- cose were similar for children and adults. For 7.5% icodextrin, Pannekeet et al. (5). A description of the calculations is given in the TCUFR in children was significantly lower than in adults Appendix A. The UFC is the product of the hydraulic permeability of Ͻ the peritoneum (L ) and the surface area (S). After calculation of the (P 0.01). Also, the marker clearance rate was significantly p lower (P Ͻ 0.02). The net ultrafiltration rate, however, was not contribution of the transcellular pores to UFCtot, it is possible to significantly different from adult patients (P ϭ 0.27). Conse- calculate the fractional transcellular UFC, meaning the part of LpS that is caused by AQP. Subsequently, it is possible to calculate the quently, the ⌬-IPV was also similar for children and adults (P fractional osmotic force exerted across the AQP. Calculations of the ϭ 0.36). fractional transcellular UFC and the fractional osmotic force across AQP were performed using the formulas described by Krediet et al. (11). A description of the calculations is given in Appendix B. Solute Transport Calculations of the fractional transcellular UFC and the fractional The transport of the low molecular weight solutes creatinine osmotic force across AQP were also performed for the adult patient and urea in children was similar for both solutions. A correla- group. tion was found between net ultrafiltration rate and MTACcreat 2942 Journal of the American Society of Nephrology J Am Soc Nephrol 15: 2940–2947, 2004

Table 1. Comparison of fluid and solute transport parameters in children and adults during a 4-hour dwell with 3.86% glucose and 7.5% icodextrina

3.86% Glucose 7.5% Icodextrin

Median Range Median Range

2 MTACurea (ml/min per 1.73 m ) child 15.4 11.9–21.2 24.9 17.6–24.9 adult 19.1 12–27 14.1 6–26 2 MTACcreat (ml/min per 1.73 m ) child 9.3 6.4–10.9 8.6 5–13.8 adult 11.7 6–21 14.3 7–24 TCUFR (ml/min per 1.73 m2) child 4.0 1.6–5.4 0.9 0.5–1.9b,c adult 4.5 0.5–6.4 2.3 1.4–4.8 MCR (ml/min per 1.73 m2) child 1.3 0.1–10.8 0.4 0–2.0d adult 1.0 0.2–4.4 1.1 0–6.8 NUFR (ml/min per 1.73 m2) child 3.0 Ϫ9.3–4.2 0.2 Ϫ1.6–1.8 adult 3.2 Ϫ1.5–5.1 1.1 Ϫ2–2.3 ⌬-IPV (ml per 1.73 m2) child 715.8 Ϫ224.2–1006.6 52.1 Ϫ394.6–443.3 adult 724 Ϫ274–1127 227 Ϫ437–527 Time on PD (mo) child 12 5.6–122.3 adult 28 3–92 D/P creatinine child 0.57 0.52–0.83 0.61 0.41–0.63

a Adult data are obtained from a previously published study (5). MTACurea, mass transfer area coefficient of urea; MTACcreat, mass transfer area coefficient of creatinine; TCUFR, transcapillary ultrafiltration rate; MCR, marker clearance rate; NUFR, net ultrafiltration rate; ⌬-IPV, intraperitoneal volume, D/P creatinine, dialysate/plasma ratio creatinine. Statistical analysis was performed with Mann- Whitney nonparametric test. b P Ͻ 0.01 3.86% glucose versus icodextrin. c P Ͻ 0.01 children versus adults. d P ϭ 0.02 children versus adults.

(r ϭ 0.69, P Ͻ 0.04) in the icodextrin dwell, which was not The mean fractional transcellular UFC calculated using found in the glucose dwell. No relation was found between 3.86% glucose was 0.15 Ϯ 0.06 ml/min per mmHg in children Ϯ Ͻ MTACcreat and age or time on PD. and 0.05 0.04 ml/min per mmHg in adults (P 0.001), No significant differences were found between children and which suggests that the AQP are responsible for, respectively, adults. A marked dip in dialysate/plasma ratio of sodium was 15 and 5% of the LpS. The fractional transcellular UFC in found in the initial phase of the 3.86% glucose dwell, which children showed a significant negative correlation with the was absent in the dwell with the icodextrin solution (Figure 2). duration of PD treatment (r ϭϪ0.67, P Ͻ 0.05) but showed no correlation with age (r ϭ 0.13, NS). In adults, there was no Contribution of the AQP to the Peritoneal UFC significant correlation with duration of treatment or age. The

The mean UFCtot calculated with 3.86% glucose solutions mean fractional osmotic force exerted across AQP using 7.5% was 0.12 Ϯ 0.002 ml/min per mmHg. The AQP contributed icodextrin was 0.18 Ϯ 0.07 ml/min per mmHg in children and 83.4 Ϯ 6.4% to this value. In the adult patient group, the AQP 0.07 Ϯ 0.06 ml/min per mmHg in adults (P Ͻ 0.001), sug- contributed 50.5 Ϯ 12% to a mean total UFC of 0.18 Ϯ 0.04 gesting that, respectively, 18 and 7% of the water flow occurs Ͻ ml/min per mmHg (P 0.01). The mean UFCaqp in children through the AQP during a dwell with icodextrin. was 0.10 Ϯ 0.02 ml/min per mmHg and in adults was 0.12 Ϯ 0.10 ml/min per mmHg (P ϭ 0.64) Discussion In both children and adults, there was no significant corre- During the past 10 to 15 years, there has been a growing lation between the UFCaqp and duration of PD treatment (chil- recognition of the need for the development of dialysis solu- dren, r ϭ 0.58, NS; adults, r ϭϪ0.56, NS) or age (children, tions, which are more biocompatible than the standard com- r ϭϪ0.12, NS; adults, r ϭ 0.19, NS). mercially available glucose-based solutions. Icodextrin is J Am Soc Nephrol 15: 2940–2947, 2004 Glucose Polymer–Based Dialysis Fluid in Children 2943

Figure 2. Dialysate/plasma ratios of sodium during 4-h dwells with 3.86% glucose (f) and 7.5% icodextrin (Œ) in children (---) and for 3.86% glucose (⅜) and 7.5% icodextrin (‚) in adults (---). The dialysate sodium concentration decreases with 3.86% glucose because aquaporin-mediated water transport exceeds the effect of sodium diffusion. Icodextrin induced no changes in dialysate/plasma sodium.

dren results in an increase of both ultrafiltration and adequacy of dialysis and that the metabolism of icodextrin occurs at a similar rate compared with that in adults (6). However, there are no data available with respect to the effect of icodextrin on the peritoneal fluid kinetics and solute transport. The present study describes the behavior of icodextrin in the peritoneal equilibration test in children and compares the results with previously published data obtained in adults (5). Figure 1. Changes of intraperitoneal volume (ml/1.73 m2) in time. The values of fluid transport parameters with 3.86% glucose Transcapillary ultrafiltration (TCUF; ᭜), marker clearance (f), and net ultrafiltration (Œ) during a 4-h dwell in children using 7.5% found in the present study were within the ranges of those icodextrin (A) and 3.86% glucose (B). TCUF after 4 h was signifi- found in a previous study performed in pediatric patients (13). cantly lower during the icodextrin peritoneal equilibration test (P Ͻ The TCUF with icodextrin was different from that obtained 0.01). Ultrafiltration and marker clearance after 4 h were similar for with 3.86% glucose. This can be explained by the fact that 3.86% glucose and icodextrin. ultrafiltration was measured after a 4-h dwell time; 3.86% glucose gives rise to a rapid ultrafiltration in the beginning of the dwell, which diminishes in time because of dissipation of mainly characterized by the absence of glucose. Instead, long- the osmotic gradient, whereas icodextrin gives rise to a slow chain glucose polymers that are responsible for ultrafiltration but sustained ultrafiltration during a prolonged period because are present. Ultrafiltration occurs according to the principle of of the limited absorption of the glucose polymers (14). colloid osmosis. The absorption of glucose polymers is limited, which gives rise to a prolonged persistence of the colloid Icodextrin in Children osmotic gradient. The use of icodextrin has been studied ex- Fluid kinetics for 3.86% glucose are similar in children and tensively in adult patients. It has been shown that the daily use adults. This is in accordance with a previous study that showed of icodextrin is safe, is generally well tolerated, and can replace that fluid kinetics in different age groups are comparable if the daily overnight use of hyperosmotic glucose solutions corrected for BSA (13,15). On the basis of these observations, (2,3). Posthuma et al. (12) demonstrated that icodextrin could it is expected that fluid kinetics for icodextrin are also similar also be very well used in an automated PD regimen to enhance between the different patient groups. Results previously pub- ultrafiltration during the long daytime dwell. The use of ico- lished by Boer et al. (6) already showed that net ultrafiltration dextrin is associated with a significant increase in serum con- obtained with icodextrin is similar for children and adults after centrations of icodextrin metabolites, which, however, has not a long-term dwell. However, our results are not fully in accor- been associated with clinical adverse events (3,12). dance with the expectations. Net ultrafiltration is similar for Until now, very little has been published about the use of both groups, but TCUF and marker clearance are significantly glucose polymer–based dialysate in children. In a previous lower in children. Rippe et al. (16) demonstrated an advantage study, it was demonstrated that the addition of a daytime of using icodextrin in patients with an increased effective icodextrin dwell to a nightly intermittent PD regimen in chil- vascular area, because icodextrin will produce an increased 2944 Journal of the American Society of Nephrology J Am Soc Nephrol 15: 2940–2947, 2004 ultrafiltration rate when the vascular surface area is increased. study group. Lai et al. (22) demonstrated that transcription and The difference in transport parameters using icodextrin be- biosynthesis of AQP-1 in human peritoneal mesothelial cells is tween adults and children thus may be explained by a differ- significantly increased upon exposure to glucose in vitro. This ence in effective vascular surface area. Although a statistical upregulation of AQP-1 upon exposure to glucose is time and difference in treatment period is not present, there seems to be dose dependent. They also demonstrated an absence of AQP-1 an overrepresentation of long-term dialysis in the adult patient in peritoneal lining denuded of mesothelial cells and speculated group. Long-term dialysis is associated with neoangiogenesis that long-term PD might lead to decreased expression of in the peritoneum (17–19), which can explain an increase in the AQP-1 on the peritoneal lining because of denudation of me- effective peritoneal surface area. However, there was no sig- sothelium. The negative correlation between the fractional nificant difference in MTACcreat, which indicates that there is transcellular UFC and treatment period in children indeed no difference in the effective peritoneal surface area. Because suggests a decreased expression of AQP in long-term PD. The of the small sample of patients, we also have to realize that absence of a relation between fractional transcellular UFC and statistical comparison is easily bothered by chance observa- age suggests that the differences observed in AQP function are tions. On the basis of these results, it therefore is not possible not related to age groups but are determined by other factors as to give an explanation for the differences seen in marker duration of treatment and glucose exposure. It is also important clearance and TCUF. As the net ultrafiltration is similar for to realize that the age range of the pediatric patients was small, children and adults, it is most likely that in clinical practice, as was the number of our observations. The current method there are no differences to be expected. makes it impossible to compare the (cumulative) glucose exposure between both study groups. A recent study showed that Osmotic Effect of Icodextrin on the Peritoneal AQP can be inactivated while they remain on the cell surface Membrane (23). This suggests that inactivation of AQP can be accom- On the basis of the three-pore model of peritoneal transport plished through means other than degradation of the water suggested by Rippe et al. (20,21), the AQP play a minor role channels. It is not yet clear which mechanism is responsible for in TCUF when using icodextrin but a major role in TCUF the inhibition of the AQP, but this might be an explanation for when using 3.86% glucose. This is demonstrated by the sig- the differences found between children and adults. It also nificant difference in contribution of AQP using 3.86% glucose should be considered that the differences in the amount of or icodextrin in both children and adults. This difference can be functional AQP are the result of a lower small pore area in visualized by analyzing the difference in sodium sieving during children. As the children are smaller, they do have lower actual the first hour of a dwell. As the water flow through the AQP small pore areas. However, by adjusting both the dwell volume will exceed the flow of water and small solutes through the and the transport parameters to the BSA, such differences small pores using 3.86% glucose, it will cause a fall in the between adults and children are no longer expected. This is sodium dialysate concentration. Using icodextrin, there will be confirmed by the fact that the MTAC for small solutes are the no fall in sodium dialysate concentration. In the present study, same in children and adults, which means that the functional this different role of the AQP is very well visualized. Using size of the small pore area will not be essentially different. 3.86% glucose, the dialysate/plasma ratio of sodium decreased, Next to the observation of differences in the amount of whereas using icodextrin, the dialysate/plasma ratio did not functional AQP, the AQP system seems also more efficient in change. The sodium dialysate/plasma curves are similar for the pediatric study group, because 83% of glucose-induced children and adults, suggesting a similar role for AQP in both ultrafiltration takes place through these AQP compared with children and adults. The theory that transport through the small 50% in adult patients. The effectiveness of the system can be pores is of great importance for the action of 7.5% icodextrin explained by the fact that LpS is a physical quantity that is is supported by the fact that in both children and adults, a defined on the basis of hydrostatic pressure (expressed as relation was found between the MTACcreat and TCUF, whereas ml/min per mmHg). The resistance caused by the ultrasmall this relation was not found using 3.86% glucose. However, our AQP is much larger than the resistance caused by the small calculated data show that a significant difference in the water pores. The reason that such a great part of the ultrafiltration flow through the AQP in both glucose- and glucose polymer– occurs through the AQP is because PD is based on a crystalloid induced ultrafiltration is present between children and adults. osmotic pressure instead of a hydrostatic pressure. This difference diminished as we calculated the absolute Further calculations show that the total amount of water, amount of the contribution of the AQP-mediated water flow to transported through the AQP during a glucose dwell (UFCaqp), the UFCtot (see below). is the same for children and adults. This suggests—although it seems that there are differences in the AQP between both study Functional Characterization of the Peritoneal groups—that the effect of the 3.86% glucose solution on the Membrane AQP is exactly the same. The UFC is the product of the peritoneal surface area (S) and It can be concluded that fluid and solute parameters are its hydraulic permeability (Lp). In children, 15% of the LpSis similar for glucose polymer–based dialysate and 3.86% glu- determined by AQP versus 5% in adults. This suggests that the cose in children, except for the TCUF. This can be explained children in our study group have a 3 times higher amount of by the absence of crystalloid osmosis and that TCUF was functional AQP as compared with the individuals in the adult measured after a 4-h dwell. J Am Soc Nephrol 15: 2940–2947, 2004 Glucose Polymer–Based Dialysis Fluid in Children 2945

Comparison of transport parameters and peritoneal mem- pores (TCUFRsp) during the initial phase of the exchange, brane characteristics reveals that there seem to be differences before absorption of solutes has taken place, equals between the peritoneal transport pathways in children and TCUFR ϭ UFC [⌬P Ϫ ͑⌸ Ϫ ⌸ ]) adults, but these differences do not interfere with the clinical sp sp c pc efficacy of the AQP because the absolute water flow through ͓ Ϫ Ϫ ϩ ͔ ϭUFCsp 9 0.38SA 7.72 0.88DIC the AQP is identical in both groups using 3.86% glucose. ϭ ͓ ϩ Ϫ Further studies are needed to explore the differences between UFCsp 1.28 0.88DIC 0.38SA] (4) children and adults in the amount and the functionality of the The TCUF during the first minute of the dwell was considered AQP and the small pores. to represent the initial TCUFR. It suggests that the UFC of icodextrin (ID) can be written as

Appendix A TCUFR0 –1min (ID) UFC ϭ ͑ml/min/mm Hg) The UFC can be calculated from the following equation: sp 0.88DIC Ϫ 0.38SA ϩ 1.28

TCUFR ϭ UFC ⅐ ͓⌬P Ϫ ␴⌬⌸ ϩ ␴⌬O] (1) (5) For 3.86% glucose, a similar equation can be given: in which TCUFR is the maximal TCUFR obtained 0–1 min ϭ ⌬ Ϫ ⌸ ϩ ␴⌬ during the first minute of an exchange, ⌬P is the hydrostatic TCUFR0–1min UFCtot[ P c O] (6) pressure gradient, ⌬͟ is the colloid osmotic pressure gradient, ⌬P was kept constant at 9 mmHg, and ͟ was calculated and ⌬O is the crystalloid osmotic pressure gradient. ⌺ is the c according to Eq. 2. ⌬O is the difference between the osmolality reflection coefficient that can range from 1.0 (ideal semiper- of the dialysis fluid (486 mOsm/L) and the plasma osmolality meable membrane) to 0 (no osmotic effect). It was assumed of the patient (Osm). As ␴ glucose averages 0.03 (27), ␴⌬O ϭ that ⌬P, during peritoneal equilibration tests, has a constant 0.03 · (486 Ϫ Osm) · 19.3 mmHg. Substitution of these value of 9 mmHg, as the capillary pressure is ~17 mmHg (24) numbers in Eq. 6 yields and the intraperitoneal pressure is 8 mmHg while resting (25). ϭ ͓ Ϫ Ϫ ϩ ͑ According to Van ‘t Hoff’s law, every mOsm/L exerts an TCUFR0–1min UFCtot 9 0.38SA 7.72 0.03 486 osmotic pressure of 19.3 mmHg in the case of an ideal semi- Ϫ ⅐ ͔ permeable membrane. This suggests that the osmotic pressure Osm) 19.3 (7) ␴ generated by an osmotic gradient is given by [osmolality · · Rearranging this equation yields 19.3]. The reflection coefficient of albumin is generally con- ͑ ͒ sidered to approach 1.0. The reflection coefficient of icodextrin TCUFR0–1min 3.86% UFC ϭ (ml/min/mm Hg) was calculated using the relation between reflection coeffi- tot 283 Ϫ 0.38SA Ϫ 0.58Osm cients of low molecular weight solutes (urea, urate, glucose, and creatinine) and albumin and their molecular weights. The (8) molecular weight of icodextrin (16,800 Da) resulted in a value Subtraction of Eq. 5 from Eq. 8 gives the UFC of the AQP: of 0.767 for the reflection coefficient. The capillary colloid ͑ ͒ osmotic pressure (͟ ) was assumed to be determined by the TCUFR0–1min 3.86% c UFC ϭ serum albumin concentration for 75% (26). To this value, 0.04 aqp 283 Ϫ 0.38SA Ϫ 0.58Osm was added because of the Gibbs-Donnan equilibrium (26): TCUFR0–1min (ID) Ϫ (9) SA ⅐ 1000 4 0.88DIC Ϫ 0.38SA ϩ 1.28 ⌸ ϭ ͫ ⅐ ϩ 0.4ͬ ⅐ 19.3 ϭ 0.38 SA ϩ 7.72 mm Hg c 68,000 3 Appendix B (2) The UFC is the product of the hydraulic permeability of the

peritoneum (Lp) and the surface area (S). It can be calculated In this equation, SA represents serum albumin (g/L), 68,000 is from the initial TCUFR and the overall peritoneal pressure the molecular weight of albumin, and the factor 1000 converts gradient according to Starling’s equation (see Appendix A): osmoles to mosmoles. The osmotic pressure within the perito- ϭ ⌬ Ϫ ␴⌬⌸ ϩ ␴⌬ ͟ TCUFR0–1min LpS[ P O] (10) neal cavity ( pc), exerted by icodextrin, equals There are apparent differences for L S values calculated using DIC ⅐ 1000 p ⌸ ϭ ͫ ͬ ⅐ ⅐ ϭ either icodextrin or glucose, whereas LpS is a membrane con- c 0.767 19.3 0.88DIC (3) 16,800 stant that is constant by definition. The most probable explanation is the heteroporosity of the peritoneum. The presence of in which DIC is the dialysate icodextrin concentration in g/L, AQP is especially important in this respect because they rep- 16,800 is the molecular weight of icodextrin, and 0.767 is the resent only a small proportion of the surface area but contribute reflection coefficient. Therefore, the TCUFR through the small largely to water flow induced by crystalloid osmosis. Despite 2946 Journal of the American Society of Nephrology J Am Soc Nephrol 15: 2940–2947, 2004

the small contribution by AQP to total peritoneal LpS, a very characteristics with glucose polymer based dialysate. Kidney Int large proportion of the osmotic force is exerted across this path- 50: 979–986, 1996 way. This is because the osmotic force is composed of the frac- 6. De Boer AW, Schröder CH, Van Vliet R, Willems JL, Monnens tional UFC values (across small pores and AQP), each multiplied LAH: Clinical experiences with icodextrin in children: Ultrafil- by the solute reflection coefficient across each pore system. tration profiles and metabolism. Pediatr Nephrol 15: 21–24, For glucose, the following calculation can be made, assum- 2000 7. Reddingius RE, Schröder CH, Willems JL, van den Brandt FCA, ing a reflection coefficient of 1.0 across the AQP and 0.03 Koomen GCM, Krediet RT, Monnens LAH: Measurement of across the small pores. peritoneal fluid handling in children on continuous ambulatory The partial osmotic forces are as follows: peritoneal dialysis using autologous haemoglobin. Perit Dial Int Aquaporins: X ⅐ L S ⅐ 1.0 (11) 14: 42–47, 1994 p 8. Nolph KD, Mactier RA, Khanna R, Twardowski ZJ, Moore H, ⅐ ϭ ͓ Ϫ McGary T: The kinetics of ultrafiltration during peritoneal dial- Small pores: (1 Ϫ X) ⅐ LpS 0.03 LpS 0.03 0.03X] ysis: The role of lymphatics. Kidney Int 32: 219–226, 1987 (12) 9. Krediet RT, Struijk DG, Koomen GCM, Arisz L: Peritoneal fluid kinetics during CAPD measured with intraperitoneal dextran 70. in which X is the part of LpS caused by AQP. ASAIO Trans 37: 662–667, 1991 The fractional osmotic force across AQP now becomes 10. Waniewski J, Werynski A, Heimbürger O, Lindholm B: Simple models for description of small solute transport in peritoneal X dialysis. Blood Purif 9: 129–141, 1991 ϭ UFCaqp (13) 0.03 ϩ 0.97X 11. Krediet RT, Lindholm B, Rippe B: Pathophysiology of peritoneal membrane failure. Perit Dial Int 20[Suppl 4]:22–42, 2000 in which UFCaqp is the contribution of AQP to UFCtot (see 12. Posthuma N, Verbrugh HA, Donker AJM, van Dorp W, Dekker Appendix A, Eq. 9). HATH, Peers EM, Oe PL, ter Wee PM: Peritoneal kinetics and A similar calculation can be made for icodextrin, assuming mesothelial markers in CCPD using icodextrin for daytime dwell a reflection coefficient of 1.0 across AQP and 0.767 across the for two years. Perit Dial Int 20: 174–180, 2000 small pores. The partial osmotic forces are as follows: 13. Reddingius RE, Schröder CH, Willems JL, Lelivelt M, Kohler BEM, Krediet RT, Monnens LAH: Measurement of peritoneal ⅐ ⅐ aquaporins: Y LpS 1.0 (14) fluid handling in children on continuous ambulatory peritoneal Nephrol Dial Transplant ͑ Ϫ ⅐ ⅐ ϭ Ϫ dialysis using dextran 70. 10: 866–870, small pores: 1 Y) LpS 0.767 LpS[0.767 0.767Y] 1995 (15) 14. Mistry CD, Gokal R, Peers E, the MIDAS Study Group: A randomized multicenter clinical trial comparing isosmolar ico- in which Y is the fractional osmotic force across AQP as dextrin with hyperosmolar glucose solutions in CAPD. Kidney calculated according to Eq. 13. Int 46: 496–503, 1994 The fractional osmotic force across AQP now becomes 15. De Boer AW, Van Schaijk TCJG, Willems JL, Reddingius RE, Monnens LAH, Schröder CH: The necessity of adjusting dialy- Y sate volume to body surface area in pediatric peritoneal equili- (16) 0.767 ϩ 0.233Y bration tests. Perit Dial Int 17: 199–202, 1997 16. Rippe B, Levin L: Computer simulations of ultrafiltration profiles for an icodextrin-based peritoneal fluid in CAPD. Kidney Int Acknowledgments 57: 2546–2556, 2000 Dr. M.M. Ho-Dac-Pannekeet is gratefully acknowledged for pro- 17. Mateijsen MA, van der Wal AC, Hendriks PM, Zweers MM, viding the data obtained in the study on adult patients. Mrs. A.J. van Mulder J, Struijk DG, Krediet RT: Vascular and interstitial Lingen-van Bueren and Mrs. T.C.J.G. van Schaijk are gratefully changes in the peritoneum of CAPD patients with peritoneal acknowledged for performing the peritoneal equilibration tests. sclerosis. Perit Dial Int 19: 517–525, 1999 18. Williams JD, Craig KJ, Topley N, Von Ruhland C, Newman GR, References Williams GT. Submesothelial fibrosis in the peritoneal mem- 1. Mistry CD, Mallick NP, Gokal R: Ultrafiltration with isosmotic brane of patients on peritoneal dialysis correlates with the pres- solution during long peritoneal dialysis exchanges. Lancet ence of vasculopathy [Abstract]. J Am Soc Nephrol 11: 314A, 2:178–182, 1987 2000 2. Mistry CD, Gokal R: Icodextrin in peritoneal dialysis: Early devel- 19. Combet S, Miyata T, Moulin P, Pouthier D, Goffin E, Devuyst opment and clinical use. Perit Dial Int 14[Suppl 2]:13–21, 1994. O: Vascular proliferation and enhanced expression of endothelial 3. Gokal R, Mistry CD, Peers E, MIDAS Study Group: A randomized nitric oxide synthase in human peritoneum exposed to long-term multicenter clinical trial comparing isosmolar Icodextrin with hy- peritoneal dialysis. J Am Soc Nephrol 11: 717–728, 2000 perosmolar solutions in CAPD. Kidney Int 46: 496–503, 1994 20. Rippe B, Stelin G, Haraldsson B: Computer simulations of perito- 4. Mistry CD, Gokal R: The use of glucose polymer (icodextrin) in neal fluid transport in CAPD. Kidney Int 40: 315–325, 1991 peritoneal dialysis: An overview. Perit Dial Int 14[Suppl 21. Rippe B: A three-pore model of peritoneal transport. Perit Dial 2]:158–161, 1994 Int 13[Suppl 1]:35–38, 1993 5. Ho-Dac-Pannekeet MM, Schouten N, Langendijk MJ, Hiralall 22. Lai KN, Li FK, Lan HY, Tang S, Tsang AWL, Chan DTM, JK, De Waart DR, Struijk DG, Krediet RT: Peritoneal transport Leung JC: Expression of aquapotin-1 in human peritoneal me- J Am Soc Nephrol 15: 2940–2947, 2004 Glucose Polymer–Based Dialysis Fluid in Children 2947

sothelial cells and its upregulation by glucose in vitro. J Am Soc 25. Imholz ALT, Koomen GCM, Struijk DG, Arisz L, Krediet RT: Nephrol 12: 1036–1045, 2001 Effect of an increased peritoneal pressure on fluid and solute 23. Jablonski EM, Webb AN, McConnell NA, Riley MC, Hughes transport during CAPD. Kidney Int 44: 1070–1085, 1993 FM Jr: Plasma membrane aquaporin activity can affect the rate of 26. Guyton AC: Textbook of Medical Physiology, 8th Ed., Philadel- apoptosis but is inhibited after apoptotic volume decrease. Am J phia, WB Saunders, 1993, pp 175–184 Physiol Cell Physiol 286: C975–C985, 2004 27. Imholz ALT, Koomen GCM, Struijk DG, Arisz L, Krediet RT: 24. Guyton AC: Textbook of Medical Physiology, 8th Ed, Philadel- Fluid and solute transport in CAPD patients using ultralow phia, WB Saunders, 1993, pp 174 sodium dialysate. Kidney Int 46: 333–340, 1995

Access to UpToDate on-line is available for additional clinical information at http://www.jasn.org/