The Effect of Increasing Hematocrit on Peritoneal Transport Kinetics1 2

Transcription

1 The Effect of Increasing Hematocrit on Peritoneal Transport Kinetics1 2 John M. Burkart,3 Barry I. Freedman, and Michael V. Rocco J.M. Burkart, B.l. Freedman. MV. Pocco. Department of Medicine, Section of Nephrology. Bowman Gray School of Medicine of Wake Forest University, Winston- Salem, NC (J. Am, Soc. Nephrol. 1994; 4: ) ABSTRACT Although it is well established that an increase in hematocrit results in a small decrease in solute transport in hemodialysis patients, the effect of hematocrit on solute transport in peritoneal dialysis patients remains controversial. In hemodialysis patients, the inverse relationship between hematocrit and the transport of small solutes may be explained by the dependence of solute clearance on the rapidity of solute movement from the red blood cell to the dialysate. This movement is determined by several in vivo factors, the most important of which is blood flow. However, in peritoneal dialysis, the effective peritoneal capillary blood flow is several times the maximal urea clearance. Therefore, clearances are usually considered to be independent of blood flow rate and variations in hematocrit should not affect solute transport across the peritoneal membrane. In this study, the effect of an increase in hematocrit on the transport of small solutes in peritoneal dialysis patients is analyzed. The peritoneal equilibration test was performed in 25 continuous ambulatory peritoneal dialysis patients before and after an increase in hematocrit of at least seven points. No significant change was found in ultrafiltration rate or in peritoneal transport characteristics as measured by the peritoneal equilibration test or by mass transfer area coefficients for creatinine, urea, or glucose. In addition, there was no change in these parameters in a, Received November 19, Accepted August 14, Portions of this work were presented at the 12th Annual Conference on Peritoneal Dialysis (February Seattle, WA) and appeared in abstract form (Peritoneal Dial nt 1992; 12(5uppl 1): 13). 3 Correspondence to Dr. J. M. Burkart. Department of Internal Medicine/Nephrology, Bowman Gray School of Medicine of Wake Forest University, Medical Center Boulevard. Winston-salem, NC I / / Journal of the American society of Nephrology Copyright C 1994 by the American Society of Nephrology control group of 13 continuous ambulatory peritoneal dialysis patients with no significant change in hematocrit. These findings are consistent with the observation that solute clearance for urea and creatinine in peritoneal dialysis is not blood flow dependent. Therefore, changes in the peritoneal transport characteristics of these solutes should not be attributed to changes in hematocrit. Key Words: Transport, hematocrit, erythropoietin, peritoneal dialysis T he introduction of recombinant human erythropoletin (rhepo) therapy has been one of the most significant recent breakthroughs in the treatment of patients with ESRD. Most dialysis patients can now be maintained at a hematocnit between 3 and 35% (1), which in turn results in a decrease in transfusion requirements along with improved cognitive function and quality of life (2-5). In hemodialysis patients. a rhepo-induced increase in med blood cell mass has been reported to significantly reduce creatinine, phosphate, and potassium clearances, but not urea clearance, during high-efficiency dialysis (6-8). This reduction in clearance occurs because the removal of solutes by hemodialysis is dependent on the flow of solute from the blood water compartment to the dialysate. If the substance being cleared has a slow med blood cell to plasma transfer mate, such as creatinine and phosphate, then a decrease in the volume of the blood water compartment will lead to a reduced clearance of that substance. If, on the other hand, there is a rapid transfer of a substance between red blood cells and plasma, as is true for urea, then a change in the blood water cornpartment volume should not appreciably affect cleanance (6,8). Thus, the reduction in the fraction of water in the blood compartment that occurs when the hematocnit Increases is responsible for the decrease in the clearance of some solutes during hemodialysis (6,9). In penitoneab dialysis patients, theme is conflicting evidence regarding the potential role of rhepo on transport characteristics. Most (1-14), but not all ( ), studies show no significant change in solute clearance or ultrafiltration mate in chronic arnbulatory peritoneab dialysis (CAPD) patients who have had an Increase in hematocnit while receiving nhepo. In pemitoneal dialysis, It is thought that effective pentoneal capillary blood flow is several times the max- I 726 Volume 4 Number

2 Burkart et al imal clearance of urea or creatinine and that the rate-limiting step in solute transport Is the resistance across the rnesothelium ( 1 9), not mesentenic blood flow. Therefore, clearances should be independent of blood flow rate and variations in hematocnit should not affect solute transport across the penitoneal membrane. In this study, using data from the pentoneal equilibration test (PET), we studied the effect of a rise in hematocnit on penitoneal solute transport and ultrafiltration rate In CAPD patients. METHODS Study Population All patients Initiating penitoneal dialysis at the Piedmont Dialysis Center since January 199 undergo PET on enrollment and at 6-mo intervals thereafter. In addition, for the purpose of this study, a PET was routinely performed before the initiation of rhepo therapy. We reviewed our patient files to identify patients on CAPD with at least two PET who met one of the following criteria: Group I-all patients with a significant increase in hernatocrit (defined as more than or equal to at least seven points) oven time with on without rhepo therapy. Group I was then subdivided into two groups on the basis of the presence (Group IA) or absence (Group IB) of rhepo therapy; Group IA-a subgroup of Group I patients who had a significant increase in hematocrit (at least seven points) since the initiation of rhepo therapy; Group lb-a subgroup of Group I patients who had a significant Increase In hematocrit (at least seven points) without receiving rhepo therapy; Group Il-patients with an increase in hematocnit of less than three points while on chronic rhepo therapy. The criteria for Groups IA and lb were chosen to allow for the companison of the effect of a substantial rise in hematocnit on an Individual patient s penitoneal membrane transport characteristics in patients receiving chronic rhepo therapy versus those not receiving rhepo therapy, respectively. Group II patients served as a control group to identify potential bong-term changes In penitoneal solute transport and ultrafiltration characteristics in patients receiving chronic rhepo therapy without a significant change in hematocrit over time. Peritoneal Transport Determinations The PET was performed by standard techniques (2). Each subject was studied after an overnight exchange of 9 to 1 2 h using the patient s standard dwell volume with 2.5% dextrose dialysate. In all cases, the dwell volume for all PET studies in an individual patient was the same. Because of the known transient effect peritonitis can have on per!- toneal membrane transport, It is standard practice at our unit not to obtain a PET until at least 1 mo after the resolution of an episode of peritonitis. The overnight dwell was drained over a 2-mm period. Then, the patient s standard dwell volume, with 2.5% dialysate solution, was infused over a 1 -mm period. During infusion, the patient was robbed from side to side every 3 mm to promote ip mixing of dialysate. A 1 -ml sample of dlalysate was removed at and 4 h for PET studies, including the measurement of creatinine, urea, and glucose. Serum chemistries and a complete blood count were obtained at the time of each PET. Dialysate and serum creatinine bevels were measured by standard automated methods by the use of a kinetic modification of the Jaffe procedure (21). Dialysate and serum creatinine values were corrected for glucose concentration by standard methods (22). Dialysate and serum glucose were measured by a modification of the hexokinase glucose-6-phosphate dehydrogenase (G-6-PD) method (23). Dialysate to plasma (D/P) ratios were calculated for creatinine and urea. Four-hour dialysate to Initial dialysate ratios for glucose (D/Do glucose) were also calculated by standard techniques (2). Mass transfer area coefficients (MTAC) for creatinine, urea, and glucose were calculated for each patient with PET data and the Pyle-Popovich formula (24). PET was delayed for at least 1 rno after the resolution of pentonitis. Erythropoietin Administration The administration of rhepo was decided on a case-by-case basis by the patient s attending physiclan. In general. patients with a hematocnit consistently less than 25% and without contmaindlcations to rhepo therapy were started on maintenance themapy. Subsequent dosage adjustments were made by the attending physician with a goal of maintaining the hernatocnit between 32 and 35%. Statistics Mean and standard deviation were calculated for all parameters. Paired T tests were used to compare PET solute transport characteristics and ultrafiltration rate observations at Time 1 and Time 2 in all four groups. Results are presented as mean ± standand deviation. Statistical significance was defined as a P value of.5 or less by a two-tailed test. RESULTS Patient Demographics Twenty-five patients had a rise In hematocrlt of at least seven points (Group I). Fourteen of these patients were receiving chronic rhepo therapy (Group IA), whereas the remaining 1 1 patients were not Journal of the American Society of Nephrology I 727

3 Hematocrit and Peritoneal Clearance (Group IB). An additional 1 3 patients had no significant change in hematocnit while on chronic rhepo therapy (Group II). Patient demographics for all four groups are listed In Table 1. The mean time interval between PET was ± I 1.2 wk in Group IA ± 18.9 wk in Group lb. and 28. ± 8.5 wk in Group II. The mean time on CAPD varied from 22.4 ± 26.3 (Group IB) to 33.2 ± 39.5 mo (Group II). Erythropoietin Dose In Group IA patients, the mean rhepo dose was for the initial PET (by definition) and 12.3 ± 5.6 U/kg body wt per week (range 33. to 29. U/kg per week) at the time of the final PET. For the patients on chronic rhepo therapy without a change In hematocnit (Group II), the mean rhepo dose at the time of the initial PET was ± 49.5 U/kg body wt per week (range, 24 to 22 U/kg body wt pen week), and It was ± U/kg body wt pen week (mange, 35.3 to 29. U/kg body wt per week) at the time of the final PET. Change in Hematocrit In Groups I. IA, and lb. there was a significant rise in hematocnit between the initial and final PET. In patients started on chronic rhepo therapy (Group IA). the hematocnit rose from a mean of 22. ± 4.4 to ± 4.38% (P <.1). In patients not receiving rhepo (Group IB), the change in hematocnit was from ± 5.39 to ± 5.74% (P <.1). For all patients with a significant rise in hematocnit (Group I), the mean hernatocnits were ± 5.2 and ± 5.1 1%, respectively (P <.1). In the control group (Group II), there was no significant change in hematocnlt (29.63 ± 4.23 versus ± 3.77%; P was not significant). Peritoneal Transport Results Group I patients did not demonstrate significant changes between initial and final ultrafiltration rate TABLE I. Patient demographics Group No. of Patients Age (yr) (mean ± SD) % Black % Women Mo on CAPD Before Stan of Study I ±14.8#{176} ±31. IA ± ±34.8 lb ± ±26.3 II ± ±39.5 /<.5 compared with Group II. and transport characteristics, including D/P creatinine, D/P urea, D/Do glucose, and MTAC for creatinine, urea, and glucose (Table 2). D/P creatinine was essentially unchanged between the first and second PET (.67 ±. 1 versus.67 ±.9, respectively), as was D/P urea (.91 ±.5 versus.91 ±.6), D/ Do glucose (.38 ±.9 versus.35 ±.7), and drain volume (2,442 ± 31 versus 2,376 ± 229 ml). Similarly, theme were no significant differences in initial and final MTAC for creatinine (12.34 ± 3.9 versus ± 2.96 ml/rnin), urea (24.1 ± 5.89 versus ± ml/min), or glucose ( ± 3.22 versus 9.58 ± 2.54 ml/min). To assess the possibility that rhepo therapy itself may cause a change in transport characteristics, Group I patients were subdivided into those receiving rhepo (Group IA) and those not receiving rhepo (Group IB). Neither of these two subgroups demonstrated any significant changes in solute transport or ultrafiltration characteristics as assessed by initial and final PET data and initial and final MTAC data (Tables 3 and 4). In the control group, who received chronic rhepo therapy without a change in hematocnit (Group II), there were no significant changes in penitoneal transport characteristics or ultrafiltration rate between the initial and final PET (Table 5). DISCUSSION In this study of transport and ultrafiltration charactenistics in CAPD patients, we have demonstrated that a rise in hematocnit with or without the use of rhepo has no significant effect on penitoneal transport properties. In addition, no changes were observed In 4-h PET drain volumes, a measure of ultrafiltration mate. In control patients who did not have a change in hematocrit, theme were also no differences in penitoneal transport characteristics. To our know!- TABLE 2. Group I-increasing hematocrit; transport characteristics Hematocrit (%) ± ± 5.llb Hemoglobin (g/dl) ± I ± I.62b 4-h D/P Creatinine.67 ±.1.67 ±.9 4-hD/PUrea.91±.5.91±.6 4-h D/Do Glucose.38 ±.9.35 ±.7 4-h Drain Volume (L) 2.44 ± ±.23 MTAC Creatinine (mi/mm) ± ± 2.96 MTAC Urea (ml/min) 24.1 ± ± 6.11 MTAC Glucose (mi/mm) 1.12 ± ± 2.54 All data are expressed as mean ± SD. b p< o.oooooi for Time I versus Time 2. For all other comparisons, P is not significant Volume 4. Number

4 Burkart et al TABLE 3. Group IA-increasing hematocrit on rhepo therapy; transport characteristics Hematocrit (%) 22. ± ± 4,38b Hemo9lobln (g/dl) 7.41 ± ± 1.39b 4-h DIP Creatlnlne.67 ±.7.66 ±.8 4-hD/PUrea.89±.6.91±.6 4-h D/Do Glucose.37 ±.7.38 ±.7 4-h Drain Volume (1) 2.42 ± ±.11 MTAC Creatinlne (mi/mm) I 1.75 ± ± 2.76 MTAC Urea (mi/mm) 23.4 ± ± 6.18 MTAC Glucose (mi/mm) 9.56 ± ± 2.66 All data are expressed as mean ± SD. b p< ooooo for Time I versus Time 2. For all other comparisons, P is not significant. TABLE 4. Group lb-increasing hematocrit without rhepo therapy; transport characteristics Hematocrit (%) ± ± 5,74b Hemoglobin (g/dl) 9.9 ± I ± I.83b 4-h D/P Creatinine.67 ± ±.1 4-hD/PUrea.92±.4.91±.5 4-h D/Do Glucose.38 ± ±.5 4-h Drain Vol (1) 2.46 ± ±.34 MTAC Creatinine (mi/mm) ± ± 3.25 MTAC Urea (mi/mm) ± ± 6.21 MTAC Glucose(mL/min) 1.98 ± ± 2.35 All data are expressed as mean ± SD. b p< o.ool for Time I versus Time 2. For all other comparisons, P is not significant. TABLE 5. Group Il-stable hematocrit on rhepo therapy; transport characteristics Hematocrit (%) ± ± 3.77 Hemoglobin (g/dl) 9.81 ± ± h DIP Creatlnine.65 ±.9.65 ±.9 4-hD/PUrea.86±.8.9±.9 4-h D/Do Glucose.4 ± ±.6 4-h Drain Volume (1) 2.64 ± ±.46 MTAC Creatlnlne (mi/mm) ± ± 3.52 MTAC Urea (mi/mm) 25.9 ± ± 9.35 MTAC Glucose (mi/mm) 9.97 ± ± 2.46 All data are expressed as mean ± SD. For all comparisons. P is not significant. edge, this study is the largest in terms of patient number and Is the only study to use a concurrent control group. The presence of a control group monitoned prospectively for the same period of time allowed us to examine if any differences in transport characteristics could be ascribed to factors other than a change in hematocnit. Nine other groups have studied the mole of changes in hematocnit on pemitoneal solute transport charactenistics. The major findings from these studies are summarized In Table 6. Hutchinson et al. (1 ) compared penitoneal transport in 14 patients with a hemogbobin of less than 8.5 g/dl with pemitoneab transport in 13 patients with a hemoglobin of more than 1.5 gjdl. There was no intergroup variation In ultrafiltratlon rate or In MTAC for urea, creatinine, or potassium. In a subgroup of eight patients who also received rhepo, theme were no differences In either ultrafiltration mate or MTAC as the hematocnit increased. Sequential studies of PET in patients with a change in hematocnit have been performed by Richmond et al. (25-27). That group has now monitored 1 2 patients over a period of 37 to 39 mo. There was a small initial decrease in D/P creatinine 1 2 to 24 wk after the start of erythropoietin therapy. D/P creatinine values then increased to a level that was 6% above baseline values at Months 25 to 27. Then, D/ P creatinine values returned to baseline by Months 35 to 37. The authors concluded that only minimal changes were seen In D/P creatinine oven the course of the study and that these changes did not interfere with the patient s dialysis regimen, consistent with our findings. They also raise the possibility that rhepo may independently influence penitoneal transport in a dose-dependent manner, a possible explanation for the disparate findings in the biteratune, but a finding not supported by our data. Taylor et at. (13), Bajo et at. (12), McMorrow and Davis (1 1), and Schmitt et at. (14) have reported on 38 CAPD patients who were monitored for up to 1 yr on nhepo therapy, and they found no significant change in solute transport or ultrafiltration rates in these patients (see Table 6). Three investigators have demonstrated that per!- toneal transport properties change as hematocnit increased. Steinhauer et at. monitored 1 4 patients with a hematocnit of less than 28% who received an initial rhepo dose of 5 U/kg body wt twice weekly. Patients were studied with 1.5-L exchanges with 1.5% glucose monohydrate dialysate. There was an increase in ultrafiltration volume from approximately 1 75 to 225 ml after rhepo therapy (17). It is unclear if the difference in ultrafiltration volume observed with 1.5-L dwell volumes would be reproducible with larger, more commonly used dwell volumes. Because other Investigators have demonstrated that net ultrafiltration decreases as dwell volume increases, we Journal of the American Society of Nephrology 1729

5 Hematocrit and Peritoneal Clearance TABLE 6. Correlation between hematocrit and peritoneal transport#{176} Author (Ref. No.) N Hematocrit (mean ± SD) Change in Transport Measurements Study Period Time I Time 2 D/P Creatinine MTAC-Creatinine Ultrafiltration (wks) Burkart et al. Studygroup ± ± Controlgroup ± ±3.7 2 Hutchinsoneto/.(1) 8 22.±3. 32.±2. N/A N/A McMorrow and Davis ± ± 2.2 N/A N/A (11) Bajoetol.(12) Tayloreto/.(13) ± ± ± ±3.5 N/A Ob SchmittetaL(14) 8 N/A N/A Ob N/A 24 Id Korbet eta!. (15) 8C 22. ± ± 3.Od Id 36 Vega et al (16) 24 N/A Hb > 1 g/dl 32 Steinhauer et a! ± ± N/A N/A I 12 (17,18) Richmond et a!. 12 N/A N/A 38 (25-27). no change in parameter;. increase in parameter;, decrease in parameter; N/A. information not reported; Hb, hemoglobin. b 24-h clearance of creatinine. CFour patients at 9 mo. d At 9 mo. hypothesize that these differences would be less significant with the more commonly used 2- and 2.5-L dwell volumes. Solute clearances of creatinine, urea, and potassium were unchanged after 1 2 mo or rhepo treatment. Phosphate clearance increased from 3.44 ±.17 to 4.38 ±.52 ml/rnin pen 1.73 m2 (18); however this change may have been due to an increase in dietary phosphorus intake ( 1 1). Kombet et at. ( 1 5) studied eight patients receiving mhepo who were monitored for variable lengths of time, up to a maximum of 9 mo. Hematocrit was more than 32% at all follow-up visits. There was no statistically significant change in creatinine clearance or D/Do glucose values at baseline compared with values obtained at 3, 6, and 9 mo after the initiation of rhepo therapy, despite an increase in hematocnit. In addition, ultrafiltration volume and MTAC creatinine were unchanged at the 3- and 6-mo follow-ups. However, D/P creatinine was significantly different from baseline values at all three follow-up periods, and the authors argue that theme was a trend for change in the other transport characteristics. Therefore, the statistically significant changes observed in penitoneal solute transport were not uniformly observed with different measures of pemitonea! transport and at all time periods examined. A mixed-effects regmession analysis demonstrated that 4-h D/P creatinine, creatinine clearance, and MTAC for creatinine decreased with increasing hematocnit levels. However, the 95% confidence limits for the slope are not meported and the use of this statistical method for small sample sizes is not we!! established (28). Using their linear regression equation for creatinine clearance in milliliters per minute, one could estimate that as the hematocnit increased from 2 to 3%, daily creatinine clearance would decrease from 1.4 to 9.6 L/ day. However, these equations were derived from a limited number of datum points (six patients at 6 mo) and assume that the average creatinine clearance in milliliters pen minute during a 4-h dwell (PET) is proportional to that obtained over a 24-h period duning which different dwell times and different dialysate concentrations are used. We have shown that estimates of creatinine clearance based on PET data are not accurate and can markedly vary from the actual measured clearances based on 24-h dialysate collections (29). Therefore, it is difficult to determine the effect of hernatocnit on 24-h clearances on the basis of the data presented by Kombet et at. (15). Vega et at. ( 1 6) reported the penitonea! clearance characteristics from a group of 24 patients studied before the initiation of rhepo and again after the hemoglobin was more than 1. g/dl. They measured 4-h D/P ratios, penitoneal clearances, and MTAC for urea and creatinine, as well as ultrafiltration volume. Statistically significant changes were observed in D/P creatinine, which decreased from.78 to.75, and in penitoneal clearance of creatinine, which decreased from 6.2 to 5.8 ml/min per square meter. There was no statistically significant change in the values for MTAC creatinine, for the ultrafiltration volume in milliliters, or for any of the measures of urea clearance, however. It is doubtful that the changes in creatinine clearance are clinically signif- I73 Volume 4 Number

6 Burkart et al icant and would require a change in dialysis pnescniption. These parameters were again measured in 15 of these patients after the rhepo dose was stable, which occurred an average of 8 mo after the start of nhepo. Unfortunately, the data from these 1 5 patients at this time interval are not compared with clearance values obtained at baseline. It is possible that rhepo therapy, independent of hematocnit, may Influence penitoneal transport and ultrafiltration mate. This hypothesis could explain the divergent results observed in prior analyses. In support of this theory, Heidenreich et at. (3) demonstrated that rhepo administration has a direct, dosedependent, vasopressom effect on proximal and mesentenic resistance vessels in the rat. This vasoconstnictive action could result in decreased mesenteric blood flow and a drop in the effective penitoneal membrane surface area and pemitoneal transport. These findings are consistent with those reported by Richmond et at. (25-27). In their study, there was a small initial drop in DIP creatinine that occurred after rhepo therapy was initiated at a dose of approximately 1 2, U/wk. However, after the hematocnit stabilized and the dose of mhepo was reduced (by an average of 6%), the D/P cneatinine increased toward baseline ( 1 3% above lowest value during peak rhepo dose). We examined whether this in vitro observation by Heidenmeich et at. has clinical relevance by comparing penitoneal transport and ultrafiltration characteristics of patients in the presence or absence of chronic rhepo therapy (Groups IA and IB, respectively). We did not demonstrate any significant difference in penitoneal solute transport on ubtrafiltration characteristics between the PET obtamed at the lower and higher hernatocnit levels in either of these two subgroups. Therefore, it is unlikely that physiologic dosages of rhepo alter mesenteric blood flow on penitoneal transport characteristics. In summary, it appears unlikely that a rise in hematocnit results in a clinically significant change in penitonea! transport properties. Oven 9 patients have now been described in the literature who have had no change in penitoneal transport characteristics despite mean increases in hematocnit of more than 1 points. The large number of patients analyzed in our study and the use of a concurrent control group address some of the deficiencies present in previous analyses. Our conclusions support the observation that the mate-limiting factor in transport across the peritoneal membrane is resistant across the penitoneal membrane, not the rate of mesenteric blood flow ( 1 9) or hematocnit. Although periodic assessment of pemitoneal transport characteristics is warranted in all penitoneal dialysis patients, changes in these characteristics should not be ascribed to changes in hematocrit or rhepo administration. ACKNOWLEDGMENTS This work was supported by a grant from the Piedmont Dialysis Nephrology Research Fund of Winston-Salem. NC. We thank Ms. Amanda Burnette for her excellent secretarial assistance. REFERENCES 1. Eschback JW, Egrie JC, Downing MR, Browne JK, Adamson JW: Correction of the anemia of end stage renal disease with recombinant human erythropoietin. Results of a combined phase I and phase II clinical trial. N Engl J Med 1987;3 16: Evans RW, Rader B, Mannigen DL: The quality of life of hemodialysis recipients treated with recombinant human erythnopoietin. JAMA 1 99;263: Eschback JW, Abduihadi MH, Browne JK: Recombinant human enythnopoietin in anemic patients with end stage renal disease. Results of a phase III multicentem clinical trial. Ann Intern Med 1989:111: Nissenson AR, Nimer SD, Wolcott DL: Recombinant human erythnopoietin and renal anemia: Molecular biology, clinical efficacy, and nervous system effects. Ann Intern Med ; 1 14: Wolcott DL, Marsh JT, La Rue A, Carr C, Nissenson AR: Recombinant human erythnopoietin treatment may Improve quality of life and cognitive function in chronic hemodiabysis patients. Am J Kidney Dis 1989:14: Shinaberger JH, Miller JH, Gardner PW: Erythropoietin alert: Risk of high hematocnit hemodiabysis. Trans Am Soc Artif Intern Organs 1988;34: Casati 5, Passerini P, Campise MR, et at.: Benefits and risks of protracted treatment with human recombinant erythnopoietin in patients having haemodialysis. BMJ l987;295: Lim VS, Flanigan MJ, Fangman J: Effect of hematocnit on solute removal during high efficiency hemodialysis. Kidney Int 199:37: Shinaberger JH, Miller JH, Gardner PW: Disadvantages and risks of normal hematocnit hemodialysis [Abstmactj. Kidney Int 1989:35: Hutchinson AJ, Ofsthun NJ, Howarth D, Gokal R: The effect of hemoglobin concentration on pemitoneal mass transfer and drain volumes in continuous ambulatory penitoneal dialysis. Peritoneal Dial Int 1 992; 12: McMorrow RG, Davis DS: The effect of an increased hematocnit on solute removal in peritoneal dialysis. Penitoneal Dial Int 1991; 11[Suppl 1J: Bajo MA, Selgas R, Miranda A, et at.: Medium term response to H-R enythropoietin In CAPD patients: The influence of erythropoietin plasma levels and the effects on penitoneab transport capacity. Adv Penitoneab Dial 1991:7: Taylor JE, MacTier PA, Henderson IS, et at.: Dialysis efficiency In continuous ambulatory penitoneal dialysis patients treated with erythropoietin. Penitoneal Dial Int 1 992; 12: Schmitt H, Schongen N, Riehl J, et at.: Eryth- Journal of the American Society of Nephrology I 731

7 Hematocrit and Perifoneal Clearance :. #{149}.. ropoietin in CAPD: Follow-up study of penitoneab membrane function IAbstnacti. Kidney mt 1992; 41: Korbet SM, Vonesh EF, Firanek CA: The effect of hematocnit on penitoneal transport. Am J Kidney Dis 1991:18: Vega N, Fernandez A, Hortal LI, et at.: Penitoneal dialysis efficiency in CAPD patients in treatment with nhuepo. Adv Penitoneal Dial 1992:8: Steinhauer HB, Lubrich-Birkner I, Dreyling KW: Increased ultrafiltration after erythnopoietin-induced correction of renal anemia in patients on continuous ambulatory penitoneal dialysis. Nephmon 1989:53: Steinhauer HB, Lubrich-Birkner I, Schollmeyer P: Effect of human recombinant erythmopoietin on dialysis efficiency in CAPD. Contnib Nephrol 1991:89: Nolph KD, Popovich RP, Ghods AJ, Twardowski Z: Determinants of low clearances of small sobutes during penitoneal dialysis. Kidney Int 1978;13: Twardowski 74, Nolph KD, Kwanna R: Penitoneal equilibration test. Pemitoneal Dial Bull 1987; 7: Cook JGH: Creatinine assay in the presence of proteinumia. Clin Chim Acta 1971:32: Cook JGH: Factors influencing the assay of creatinine. Ann Clin Biochem 1975:12: Stein MW. In: D-Gbucose determination with hexokinase and g!ucose-6-phosphate dehydrogenase. Bengmeyen HU, Ed. Methods of Enzymatic Analysis. New York: Academic Press; 1965: Popovich RP, Pyle WK, Moncrief JW. Kinetics of penitoneal transport. In: Nolph DK, Ed. Pentoneal Dialysis. Boston: Martinus Nijhoff; 1981: Richmond D, Broyan P, Shea 5, Reft C, Poseno M, Gimenez LF: How does rhuepo effect D/P creatinine ratios? Adv Penitonea! Dial 1992:8: Richmond D, Reft C, Poseno M, Shea 5, Broyan P: What will rhuepo do to ultrafiltration [Abstracti? Penitoneal Dial Int ; 1 1 [Suppl 1]: Richmond DJS, Broyan P. Poseno M, Reft C, Shea 5, Gimenez L: Long term rhuepo therapy effects on D/P cneatinine ratios [Abstract]. Pentoneal Dial Int 1 993; 1 3[Supp! 11: Vonesh EF: Efficient inference for random-coefficient growth curve models with unbalanced data. Biometrics 1987:43: Burkart JM, Jordan JR, Rocco MV: Assessment of dialysis dose by measured clearance versus extrapolated data. Penitoneal Dial Int 1993;13: Heidenreich 5, Rahn KH, Zidek W: Direct vasopnessor effect of recombinant human emythnopoietin on renal resistance vessels. Kidney Int 1991:39: I 732 Volume 4 Number