Middle molecule and small protein removal in children on peritoneal dialysis

Kidney International, Vol. 61 (2002), pp. 1153 1159 Middle molecule and small protein removal in children on peritoneal dialysis GIOVANNI MONTINI, GIANPAOLO AMICI, SABRINA MILAN, 1 MICHELE MUSSAP, MAURO NATURALE, ILSE-MARIA RÄTSCH, ANITA AMMENTI, PALMA SORINO, ENRICO VERRINA, BARBARA ANDREETTA, and GRAZIELLA ZACCHELLO, on behalf of the ITALIAN REGISTRY OF PEDIATRIC PERITONEAL DIALYSIS 2 Nephrology, Dialysis and Transplant Unit, Pediatric Department, Padova; Nephrology, Dialysis and Transplant Unit, Treviso; Department of Laboratory Medicine, Padova; Nephrology and Dialysis Unit, Pediatric Department, Parma; Nephrology and Dialysis Unit, Pediatric Department, Ancona; Nephrology and Dialysis Unit, Bari; and Nephrology, Dialysis and Transplant Unit, Genova, Italy Middle molecule and small protein removal in children on perifor 1.73 m 2 ) and weekly solute removal index (SRI) were calculated toneal dialysis. all the solutes; weekly Kt/V was calculated for urea. Background. Dialysis efficiency has a great influence on the Results. In non-anuric versus anuric patients the total clearoutcome of patients. Few data are available on the removal of ances were: urea 82.6 18.3 versus 71.3 26.4; creatinine 82.7 solutes with molecular weights higher than urea and creatinine. 28.6 versus 47.8 18.8; inulin 42.8 11.3 versus 32.8 20.4; The aim of our study was to assess the transport and the re- 2 m 14.2 13.8 versus 9.2 8.3; cystatin C 20.2 9.4 versus moval of substances with molecular weights up to 15 kd and 9.7 4.8. In the patients with residual diuresis, the urea was to evaluate the contribution of residual renal function in perito- removed mainly by PD (69.2%), while inulin, 2 m and cystatin neal dialysis (PD) children. C were removed by renal clearance (64.0%, 79.5% and 62.8%, Methods. Seventeen patients of 12 4 years undergoing autoin respectively). Total, peritoneal and renal weekly Kt/V values mated PD were studied. Ten patients had 563 355 ml/ the subjects with residual renal function, were 2.86 0.70, day of urine output, and 7 were anuric. During a standardized 1.99 0.40 and 0.87 0.43, respectively. Peritoneal weekly nightly intermittent PD (NIPD) session, a single-injection inuin Kt/V in the anuric patients was 2.36 0.85; total weekly Kt/V lin clearance was performed. Urea, creatinine, inulin (meaversus the total group was 2.65 0.78. Weekly SRIs in non-anuric sured by HPLC), cystatin C and 2 -microglobulin ( 2 m) were anuric patients were: urea 2.56 0.58 versus 2.09 measured in blood, urine and dialysate. Clearances (L/week/ 0.74; creatinine 2.66 0.73 versus 1.46 0.56; inulin 2.36 0.92 versus 1.64 1.60; 2 m 1.26 1.10 versus 1.20 1.90; cystatin C 1.72 0.83 versus 1.58 1.62. 1 S.M. is presently at the Nephrology Unit of Castelfranco. Conclusions. Solutes removed during PD tend to decrease fol- 2 Investigators and their affiliated institutions participating in the Italian lowing an increase in molecular weight of the substance. Since Registry of Pediatric Peritoneal Dialysis include: I. Rätsch (Nephrology anuric patients are at higher risk of middle molecule and small and dialysis unit, Ancona); D.A. Caringella, P. Sorino (Nephrology and protein accumulation, more attention should be paid to the dialysis unit, Giovanni XXIII Hospital, Bari); L. Catizone, (Nephrology removal of middle molecules. Further studies should be underand dialysis unit, Malpighi Hospital, Bologna); G. Cancarini (Nephrol- taken to evaluate whether removing them has a clinical impact ogy and dialysis unit, Brescia); G. Lavoratti (Pediatric department, Meyer and to determine their threshold levels. Hospital, Firenze); F. Perfumo, E. Verrina (Nephrology, dialysis and transplant unit, Istituto G. Gaslini, Genova); A. Edefonti (Nephrology, dialysis and transplant unit, Department of pediatrics, De Marchi, Milano); C. Pecoraro (Nephrology and dialysis unit, Santobono Hospital, Napoli); G. Zacchello, B. Andreetta (Nephrology dialysis and transplant Uremic toxicity closely depends on the accumulation unit, Università, Padova); S. Maringhini 3 (Nephrology and dialysis unit, of many substances, including not only small plasma Di Cristina Hospital, Palermo); G. Rizzoni, S. Rinaldi (Nephrology di- solutes, but also the so-called middle molecules, with alysis and transplant unit, Ospedale Bambino Gesù, Roma); R. Coppo, B. Gianoglio (Nephrology dialysis and transplant unit, Ospedale Regina a molecular weight (MW) ranging from 0.3 to 5 kd, and Margherita, Torino). low-molecular mass proteins (MW 40 kd) [1]. Wellknown examples are parathyroid hormone (MW 9.5 kd), Key words: automated peritoneal dialysis, dialysate removal, cystatin C, 2 -microglobulin, inulin, clearance, Kt/V, solute removal index, mass 2 -microglobulin ( 2 m; MW 11.8 kd) and leptin (MW transfer. 16 kd) [2, 3]. In adult studies there has been increasing recognition Received for publication May 16, 2001 and in revised form October 18, 2001 of the influence of adequate solute clearance on patient Accepted for publication October 22, 2001 outcome in peritoneal dialysis (PD). Both clinical (feeling 2002 by the International Society of Nephrology of well-being and the absence of uremic symptoms) 1153

1154 Montini et al: Middle molecule removal in dialysis children Fig. 1. Experimental protocol. and biochemical (urea and creatinine clearance, and weekly Kt/V for urea) parameters are used to assess the adequacy of PD, even if there are limited data from intermittent PD therapies [4]. The assessment of dialysis adequacy in children is even more complicated, since patient size varies widely [5]. Furthermore, dialysis efficiency has significant effects on statural growth and nutritional parameters: Schaefer, Klaus and Mehls have demonstrated that the peritoneal transporter state is a weak, but significant determinant of growth and body weight gain in children [6]. Recently Verrina et al have proposed the solute removal index (SRI) as an index of dialysis efficiency focused on urea mass balance with specific advantages in the comparison of different dialysis techniques [7]. Very few data are available on the clearance of solutes with higher MW, which can contribute to uremic symptoms (for example, decreased appetite) and to the longterm outcome of the patient maintained by PD. Residual renal function has a major role: even relatively small residual urea clearances may contribute substantially to overall clearance among PD patients. Inulin (5.5 kd) already has been utilized as a marker of the clearance of middle molecules in adult hemodialyzed patients. Cystatin C, a low-mw plasma protein (13 kd) consisting of 120 amino acid residues, has been recently proposed as a very sensitive endogenous marker of changes in glomerular filtration rate (GFR). This protein might reflect the dialytic efficiency of PD in removing low-mw proteins from plasma. We have performed this study to assess the transport and peritoneal removal of substances with a MW up to 15 kd, and to evaluate the contribution of residual renal function in pediatric patients maintained by automatic nightly peritoneal dialysis. METHODS Patients Seventeen children (5 males) from five pediatric nephrology centers, with a mean age of 12 4 years (range, 5 to 18) undergoing automated PD (APD) were studied. Median duration of dialysis at time of enrollment was 15 months (range, 3 to 122), pre-dialysis body weight was 31.4 10.2 kg (range, 12.0 to 53.5) and body surface area (BSA) was 1.07 0.25 m 2. At the time of the study, none had any condition, such as chronic liver or lung dis- ease, malignancy or inflammatory disorders, likely to interfere with protein metabolism. All patients had been free from peritonitis for at least one month prior to the study. Ten patients had 563 355 ml of urine output (median 475, range, 100 to 1200), 7 were anuric (no patient was anephric). Informed consent was obtained from par- ents and/or patients, as appropriate. Study protocol All patients were dialyzed with an automated, nightly intermittent PD (NIPD) standardized protocol: 12 hours, 12 exchanges with dwell times of 45 minutes, fill volume 1100 ml/m 2 of body surface area, 1.36% glucose solution. One hour before starting NIPD (Fig. 1), inulin (Jacopo Monico, Venezia, Italy) was administered as a single in- jection (103 19 mg/kg, range from 94 to 175) within two minutes. Serum samples were withdrawn after 15, 40 and 60 minutes from the injection of inulin, to evaluate its vol- ume of distribution, and then NIPD was started. Blood, peritoneal dialysate and urine samples were taken as follows. At 0 minutes and 12 hours, blood (2 ml) for urea, creatinine, inulin, cystatin C, 2 m was taken; blood (1 ml) for inulin was taken at 15, 40, and 60, minutes, and 4 and 8 hours. The total amounts of urine and peritoneal drainage were collected, weighed and sampled (10 ml) for urea, creatinine, inulin, cystatin C, and 2 m assays. At the second NIPD cycle, 10 ml of the drained dialysate also were taken for the same assays. Analytical methods Inulin was measured by a previously described, rapid high-performance liquid chromatography (HPLC) method, utilizing ion moderated partition chromatography and evaporative light-scattering detection [8]. The

Montini et al: Middle molecule removal in dialysis children 1155 equipment consists of a liquid chromatograph containing Table 1. Pre- and post-dialysis biochemical parameters of the 17 patients studied the two-pumps model 510 (Waters Assoc., Milford, MA, USA), a detector Model 450, an autosampler Model 465, Plasma values Pre-dialysis Post-dialysis t test P and a chromatography, data system Model 450-MT (Kon- Urea mmol/l 28.5 8.7 23.6 7.8 0.001 tron Instruments, Milan, Italy). Cystatin C was determined Creatinine lmol/l 894 233 815 218 0.001 Inulin mg/ml 0.56 0.41 0.39 0.40 0.001 by automated PETIA method (Dako, Milano, Italy; basic 2-microglobulin mg/l 28.5 11.5 31.3 11.9 0.031 a measuring range covers cystatin-c concentrations from Cystatin-C mg/l 6.9 1.6 7.5 1.4 0.002 a 0.40 to about 14.1 mg/l) [9], 2 m by the immunonephelometric a Post-values are more than pre-values method (Dade, Milano, Italy), and creatinine and urea were measured by enzymatic assay (Ortho Clinical Diagnostics, Milano, Italy). Statistical analysis Calculations Data are expressed as mean standard deviation. Differences Inulin single-shot clearance was calculated from the were evaluated by the paired and unpaired Stu- plasma inulin disappearance rate as the injected dose/ dent t test, where appropriate. Pearson correlation co- total area-under-the-plasma-curve (AUC) [10], and from efficients were calculated to assess possible associations dialysate and urine inulin appearance rate of 12 hours between clearance and plasma values of the different soas: mass transfer lutes. Normality of the distributions was assessed by Sha- 0 12 /AUC 0 12. The least-squares iterative computerized fitting according to Gauss-Newton-Raphtests with two-tailed alpha values lower than 0.05. JMP piro-wilk test. The null hypothesis was rejected for all son [11] was carried out on inulin plasma values using the standard biexponential model Aexp(-at) Bexp(-bt) 3.2.2 (SAS, Cary, NC, USA) and Pro Fit 5.1.1 (Quan- where t is time and A, a, B, b are non-linear fitting cothe statistical analysis and for the iterative computer- tum Soft, Zurich, CH) software programs were used for efficients. The total and 12-hour AUCs were calculated using fitting parameters by the formulas: AUC A/a ized fitting. B/b and AUC 0 12 A/a[exp(-at 0 ) exp(-at 12 )] B/b [exp(-bt 0 ) exp(-bt 12 )] [12]. The apparent distribution RESULTS space of inulin was calculated by the injected dose/ Pre- and post-dialysis blood levels of urea, creatinine, bauc, and was taken as extracellular volume measure inulin, 2 m and cystatin C of the 17 patients are shown (ECV, liters) [13]. in Table 1. Urea, creatinine and inulin plasma levels Clearances (L/week) were calculated for all solutes by were significantly lower after dialysis treatment; the two direct determination from dialysis fluids, urine and mean higher MW compounds 2 m and cystatin C showed an pre- and post-dialysis serum solute concentrations, based opposite behavior, significantly increasing their plasma on a single day collection. Weekly Kt/V was calculated levels as a consequence of dialysis ultrafiltration (UF). only for urea by direct determination from dialysis fluids, UF was similar in the anuric and non-anuric children urine and mean pre- and post-dialysis serum urea concen- (372 333 vs. 374 167 ml). trations. Total body water (liters) was assumed as urea Plasma values of the following substances were sigdistribution space [14]. nificantly higher in anuric versus non-anuric children: cre- Peritoneal mass transfer area coefficients (MTAC, atinine (pre-dialysis 1119.1 177.8 vs. 736.4 89.6 mol/l, ml/min) were calculated from data obtained at the sec- P 0.001), cystatin C (pre-dialysis 7.9 1.7 vs. 6.2 ond cycle of the NIPD session, as previously suggested 0.5 mg/l, P 0.037) and 2 m (pre-dialysis 42.6 15.5 for adult patients by Garred, Canaud and Farrell [15] and vs. 23.2 4.0 mg/l, P 0.008); anuria did not influence adapted to children by Geary, Harvey and Balfe [16]. urea plasma levels significantly (pre-dialysis 28.0 9.7 vs. Normalizations for MTAC and clearances were calcu- 28.8 8.4 mmol/l, P NS). As a consequence of the lated by 1.73 m 2 of body surface area. intravenous bolus, pre-dialysis plasma levels of inulin Mass transfer (MT) was calculated for all solutes by were not different; instead, post-dialysis plasma levels direct determination from dialysis fluids and urine with- were significantly higher in anurics than in non-anurics out applying any normalization or correction. (0.35 0.1 vs. 0.26 0.5 mg/ml, P 0.020). Weekly SRI was calculated for all solutes as the ratio Bi-exponential disappearance single-shot inulin curves between mass transfer and predialysis body content [7, are shown in Figure 2. The half-life of the inulin distribu- 17, 18]. Predialysis body content was calculated for each tion phase was 27 20 minutes, the half-life of the elimisolute by multiplying distribution space for predialysis nation phase was 29 17 hours, and the 95% plasma serum concentrations. The distribution space was assumed elimination was 5.2 3.1 days. The extracellular volume, to be total body water for urea and creatinine and extracel- evaluated as inulin distribution space, was 6.6 2.8 L and lular volume for inulin, 2 m and cystatin C [2, 17 20]. represented 22.0 7.1% of body weight. A statistically

1156 Montini et al: Middle molecule removal in dialysis children Fig. 2. Bi-exponential disappearance single shot inulin curves. Inulin (100 mg/kg body weight) was injected at baseline (t 0 minutes). Each line corresponds to a single plasma clearance of inulin. The fast disappearance phase corresponds to the distribution phase and the slow disappearance to the real inulin clearance from the body. Table 2. Peritoneal, renal, and total clearances of various solutes with different molecular weights (MW) Urea Creatinine Inulin 2 -Microglobulin Cystatin C MW Daltons 60 113 5500 11800 13200 All patients (N 17) Total clearance L/week/1.73 m 2 77.9 22 68.4 28.6 38.7 15.9 12.2 11.8 16.0 9.3 Patients with maintained diuresis (N 10) Peritoneal clearance L/week/1.73 m 2 57.2 8.7 38.0 14.2 15.4 4.5 2.9 2.1 7.4 1.2 Peritoneal clearance % of total 69.2% 45.9% 35.9% 20.4% 36.6% Renal clearance L/week/1.73 m 2 25.3 13 44.7 25.6 27.4 14.2 11.3 13.6 12.7 8.9 Renal clearance % of total 30.6% 54.0% 64.0% 79.5% 62.8% Total clearance L/week/1.73 m 2 82.6 18.3 82.7 28.6 42.8 11.3 14.2 13.8 20.2 9.4 Anuric patients (N 7) Peritoneal clearance L/week/1.73 m 2 71.3 26.4 47.8 18.8 32.8 20.4 9.2 8.3 9.7 4.8 Fig. 3. Relationship of peritoneal mass transfer area coefficients (MTAC) to molecular weights. There is an inverse non-linear relationship be- tween the molecular weight and the peritoneal diffusive capacity. Molecular weight (MW) is expressed in daltons. significant difference was observed when comparing inulin bi-exponential disappearance clearance (54.1 27.0) and total clearance (38.7 15.9 L/week/1.73 m 2, P 0.026) obtained from urinary and peritoneal inulin appearance. Total and separated (dialytic and urinary) clearance values for urea, creatinine, inulin, cystatin C and 2 m are summarized in Table 2. Data of clearances are also expressed as the percentage (dialytic vs. renal) of the total values in the ten children who maintained residual renal function, showing progressive decrease of the dialytic percentage with increase in the MW. Similarly, peritoneal MTAC, normalized for 1.73 m 2 of body surface, showed a decreased slope that was inversely proportional to the MW of the substance, as a hyperbolic function (Fig. 3). Total clearances were significantly higher in non-anuric versus anuric children for creatinine (P 0.008) and for cystatin C (P 0.026); no differences were observed in urea, inulin and 2 m beta-2-microglobulin (Table 2).

Montini et al: Middle molecule removal in dialysis children 1157 Table 3. Peritoneal, renal, and total mass transfer (MT) of various solutes with different molecular weights (MW) Urea Creatinine Inulin 2 -Microglobulin Cystatin-C MW Daltons 60 113 5500 11800 13200 All patients (N 17) Total MT mg 4719 1580 556 251 1003 987 27 23 10 5 Patients with maintained diuresis (N 10) Peritoneal MT mg 3409 981 246 78 204 116 6 4 4 1 Renal MT mg 1628 1044 312 214 757 519 18 18 7 5 Total MT mg 5037 1888 558 228 961 519 25 17 11 6 Anuric patients (N 7) Peritoneal MT mg 4264 947 554 301 1063 1479 29 30 7 3 All values are expressed as absolute solute output in mg of a single automated peritoneal dialysis (APD) session without any normalization or correction. Table 4. Peritoneal, renal, and total solute removal index (SRI) of various solutes with different molecular weights (MW) Urea Creatinine Inulin 2 -Microglobulin Cystatin-C MW Daltons 60 113 5500 11800 13200 Distribution compartment TBW TBW ECV ECV ECV All patients (N 17) Total SRI ( 100) 2.37 0.67 2.16 0.88 2.06 1.25 1.23 1.45 1.67 1.16 Patients with maintained diuresis (N 10) Peritoneal SRI 100 1.78 0.32 1.23 0.47 0.51 0.26 0.26 0.18 0.63 0.14 Renal SRI 100 0.78 0.38 1.42 0.73 1.86 1.04 1.00 1.09 1.10 0.76 Total SRI 100 2.56 0.58 2.66 0.73 2.36 0.92 1.26 1.10 1.72 0.83 Anuric patients (N 7) Peritoneal SRI 100 2.09 0.74 1.46 0.56 1.64 1.60 1.20 1.90 1.58 1.62 All values are expressed on a weekly basis and in percentage of pre-dialysis body content. The distribution compartments are assumed to be total body water (TBW) 18.7 5.4 L (60.4 4.3% of body weight) and extracellular volume (ECV) 6.6 2.8 L (22.0 7.1% of body weight) for different solutes. SRI is the solute removal index. Total, peritoneal and renal urea weekly Kt/V in the quacy is mandatory, especially for this percentage of 10 children who maintained residual renal function was young patients. 2.86 0.70, 1.99 0.40 and 0.87 0.43, respectively. Adult studies [22] have suggested that higher clearances Peritoneal urea weekly Kt/V in the 7 anuric patients was of small molecules are associated with better sur- 2.36 0.85, total urea weekly Kt/V in the total group vival and lower morbidity and [23] have set the weekly of 17 children was 2.65 0.78. MT and SRIs are shown targets at 2.0 for Kt/V and 60 L/1.73 m 2 for creatinine in Tables 3 and 4. clearance, with an increase of 10% for NIPD. Similar or Plasma cystatin C values showed a significant inverse greater targets have been extended to pediatric patients correlation with renal inulin clearance (r 0.615; P [24 26]. 0.008) and total creatinine clearance (r 0.558; P Recently markedly elevated leptin (a low-mw plasma 0.020). protein, product of fat cells, which regulates food intake, The ratio between the renal clearance of inulin and the through modulation of satiety signals, and energy expenother four solutes showed mild tubular reabsorption of diture in animal models) levels have been documented in urea (0.96 0.30) and a much more consistent reabsorpdialysis; uremic patients, especially in those treated by peritoneal tion of 2 m (0.43 0.44) and cystatin C (0.47 0.20); tuchanges this may contribute to uremic anorexia and bular secretion of creatinine was observed (1.80 1.35). in body composition [27]. This low-mw plasma protein could be one of the many substances, which DISCUSSION accumulate in a uremic organism. In fact, normal kidneys not only excrete water, electrolytes, and small molecular Data from the Italian Registry of Pediatric Chronic end-products of metabolism, but they also have a role Peritoneal Dialysis shows that 297 children were main- in removing larger molecules: peptides and low-mw protained by CPD during the years 1986 to 1997 [21]. CPD teins [1]. patients represented an increasing percentage of all chil- In the clinical setting blood concentrations of low-mw dren on chronic dialysis in the Italian Pediatric Dialysis substances are utilized to monitor dialysis efficiency, and Centers: from 42.6% in 1986, to 64.5% in 1997. Nine little attention is paid to compounds with higher MW percent of the children remained on CPD longer than [28]. This aspect could be particularly important in patients five years; therefore, an effort to optimize dialysis ade- with no residual diuresis, who depend completely

1158 Montini et al: Middle molecule removal in dialysis children on dialysis efficiency. Recent studies in adult hemodialy- an alternative efficiency dialytic index, to be utilized as a sis patients show that the use of high flux membranes unified basis for comparing different dialysis techniques removing more middle molecules ensures an excellent [7, 17, 18]. We have calculated SRIs values (Table 4) dialysis quality, influencing the clinical outcome of patients not only for urea, but also for creatinine, inulin, 2 m and in several areas ( 2 m associated amyloidosis, re- cystatin C. The distribution space was assumed to be covery from acute renal failure and mortality of chronic total body water for urea and creatinine, and extracellu- hemodialysis patients) [29]. Indeed, we have shown that lar volume for inulin, 2 m and cystatin C. SRI values for residual renal function has a central role in removing urea are similar to the values reported by Verrina et al some low-mw plasma proteins and larger middle mole- [7]. To the best of our knowledge, there are no SRI values cules. Urea is mainly removed from the plasma by peritoneal reported for molecules with a higher MW. It is intercules. dialysis (69.2%); inulin, 2 m and cystatin C by esting to note that the percentage of the removal index renal clearance (64.0%, 79.5% and 62.8%, respectively) is distributed in a similar proportion to the respective in the patients with urine output (Table 2). Therefore, clearances, between the kidney and the peritoneum for patients with no diuresis remove higher MW substances the different molecules. If clearances decrease following with greater difficulty. the increase in MW in a scale from 77.9 (creatinine) to Urea weekly Kt/V values, as well as peritoneal and renal 12.2 ( 2 m) L/week/1.73 m 2, SRI decreases from 2.37 to clearances, showed satisfactory values when com- 1.23, respectively. This means that normalizing removal pared to the recommended targets [22 26] in both anuric by the solute content of the body at the beginning of and non-anuric patients. As recently reported in a study the treatment, and therefore taking into consideration on APD pediatric patients [30], the recommended targets the distribution space, could allow a more useful compar- for creatinine clearance are more difficult to meet, ison of middle molecules and small protein removal than at least by peritoneal dialysis alone; in our experience, normalizing by body surface area. mean creatinine clearance was unsatisfactory in anuric In conclusion, the evaluation of PD efficiency may be children (47.8 L/week/1.73 m 2 vs. the recommended value determined more effectively by taking into account the of 60 in adults) and very satisfactory in children with a clearance not only of the standard small solutes, but preserved diuresis (83 L/week/1.73 m 2 ). The central role also of middle molecules and low-mw proteins, which of the residual diuresis has been previously shown in better reflect the complexity of uremia. Peritoneal clearadult patients [31]: it emerged as the most important de- ances and SRIs tend to decrease following the increase terminant of serum low-mw protein concentration; trans- in MW of the substance, and removal of the middle moleperitoneal clearance was dependent mainly on the size cules is superior when residual renal function is mainof the protein. The dwell time emerged as an important tained. Therefore, more attention should be paid to the factor only for high ( 60 kd) MW proteins. In our pe- removal of middle molecules. Further studies should be diatric patients the MTAC showed a similar result for undertaken to evaluate whether removing them has a small MW proteins. Therefore, it is clear that 1 ml/min of clinical impact and to determine threshold levels. inulin renal clearance is different from 1 ml/min of inulin peritoneal clearance, because there is progressively ACKNOWLEDGMENTS increasing peritoneal resistance to the transport of sub- This paper was presented in part at the European Society of Pedistances with increasing MW, especially when it is greater atric Nephrology (2000) and at the American Society of Nephrology than 5 kd (Table 2). Furthermore, even in end-stage renal (1999). failure, the kidney has an active tubular role in creati- Reprint requests to Giovanni Montini, M.D., Nephrology, Dialysis nine secretion, and 2 m and cystatin C reabsorption. The and Transplant Unit, Pediatric Department, University Hospital, Via peritoneal clearance of cystatin C appears to be greater Giustiniani, 3, 35128 Padova, Italy. than that of 2 m, despite the fact that cystatin C has the E-mail: montini@pediatria.unipd.it largest MW and the smallest MTAC. This could be partly due to the high variability of 2 m, and an influence of the REFERENCES shape and molecular radius of the proteins. 1. Bergström J, Whele B: Clinical implication of middle and larger molecules, in The Kidney (5 th ed, vol 26), edited by Brenner BM, Our data also show that plasma values of creatinine, Philadelphia, Saunders, 1996, pp 579 590 cystatin C and 2 m were significantly higher in anuric than 2. Vanholder R, De Smet R: Pathophysiologic effects of uremic in non-anuric children. Renal diuresis is inversely corre- retention solutes. J Am Soc Nephrol 10:1815 1823, 1999 lated with serum cystatin C levels; therefore, patients withtion of serum leptin levels in children with chronic renal failure. 3. Daschner M, Tonshoff B, Blum WF, et al: Inappropriate eleva- out residual renal function have higher plasma levels of European Study Group for Nutritional Treatment of Chronic Renal Failure in Childhood. J Am Soc Nephrol 9:1074 1079, 1998 cystatin C than patients with residual diuresis, even when the usual biochemical parameters of dialysis efficiency 4. Burkart JM, Henrich WL: Adequacy of continuous peritoneal dialysis, in Uptodate in Nephrology and Hypertension, July 1999 are satisfactory. http://www.uptodate.com It has been previously suggested that SRI represents 5. Balfe JW: Peritoneal dialysis, in Pediatric Nephrology (2nd), ed-

Montini et al: Middle molecule removal in dialysis children 1159 ited by Holliday MA, Barratt TM, Vernier RC, Baltimore, Wil- nutritional status in haemodialysis patients. Nephrol Dial Transplant liam & Wilkins, 1987, pp 814 822. 10:1405 1410, 1995 6. Schaefer F, Klaus G, Mehls O: Peritoneal transport properties 20. Guyton AC: Textbook of Medical Physiology (8th ed). Philadelphia, and dialysis dose affect growth and nutritional status in children WB Saunders, 1991, pp 274 307 on chronic peritoneal dialysis. Mid-European Pediatric Peritoneal 21. Verrina E, Perfumo F, Calevo MG, et al: The Italian Pediatric Dialysis Study Group. J Am Soc Nephrol 10:1786 1792, 1999 Chronic Peritoneal Dialysis Registry. Perit Dial Int 19(Suppl 2): 7. Verrina E, Brendolan A, Gusmano R, Ronco C: Chronic renal S479 S483, 1999 replacement therapy in children: Which index is best for adequacy? 22. Canada USA (CANUSA) Peritoneal Dialysis Study Group: Kidney Int 54:1690 1696, 1998 Adequacy of dialysis and nutrition in continuous peritoneal dialysis: 8. Marsilio R, Naturale M, Manghi P, et al: Rapid and simple Association with clinical outcomes. J Am Soc Nephrol 7:198 determination of inulin in biological fluids by high-performance 207, 1996 liquid chromatography with light-scattering detection. J Chromatogr 23. Golper TA, Twardowski ZJ, Warady BA, et al: Dose and ade- B Biomed Sci Appl 744:241 247, 2000 quacy. Perit Dial Int 17(Suppl 3):S40 S41, 1997 9. Kyhse Andersen J, Schmidt C, Nordin G, et al: Serum cystatin 24. NKF-DOQI Clinical Practice Guidelines for Peritoneal Dialysis C, determined by a rapid, Automated Particle-Enhanced Turbi- Adequacy. National Kidney Foundation. Am J Kidney Dis 30 dimetric method, is a better marker than serum creatinine for (Suppl 2):S67 S136, 1997 glomerular filtration rate. Clin Chem 40:1921 1926, 1994 25. Holtta T, Ronnholm K, Jalanko H, Holmberg C: Clinical outcome 10. Florijn KW, Barendregt JN, Lentjes EG, et al: Glomerular filtration of pediatric patients on peritoneal dialysis under adequacy rate measurement by single-shot injection of inulin. Kid- control. Pediatr Nephrol 14:889 897, 2000 ney Int 46:252 259, 1994 26. Schaefer F, Wolf S, Klaus G, et al: Higher KT/V urea associ- 11. Draper NR, Smith H: Applied Regression Analysis (2 nd ed). New ated with greater protein catabolic rate and dietary protein intake York, Wiley, pp 458 517, 1981. in children treated with CCPD compared to CAPD. Mid-Euro- 12. Bianchi C: Measurement of the glomerular filtration rate, in Prog- pean Pediatric CPD Study Group (MPCS). Adv Perit Dial 10:310 ress in Nuclear Medicine (vol 2), Basel, Karger, 1971, pp 21 53 314, 1994 13. Gibaldi M, Perrier D: Pharmacokinetics (2 nd ed). New York, Dek- 27. Stenvinkel P, Lindholm B, Lonnqvist F, et al: Increases in serum ker, 1982, pp 199 219 leptin levels during peritoneal dialysis are associated with inflam- 14. Mellits ED, Cheek DB: The assessment of body water and fat- mation and a decrease in lean body mass. J Am Soc Nephrol 11: ness from infancy to adulthood. Monogr Soc Res Child Dev 35: 1303 1309, 2000 12 26, 1970 28. Laux C, Weiss B, Bonzel KE: Middle molecules in peritoneal 15. Garred LJ, Canaud B, Farrell PC: A simple kinetic model for equilibration test as a marker of peritoneal stress in children on assessing peritoneal mass transfer in chronic ambulatory peritoneal continuous peritoneal dialysis. Adv Perit Dial 15:283 286, 1999 dialysis. ASAIO J 6:131 137, 1983 29. Cheung AK, Leypoldt JK: The hemodialysis membranes: A his- 16. Geary DF, Harvey EA, Balfe JW: Mass transfer area coefficients torical perspective, current state and future prospect. Semin Nephrol in children. Perit Dial Int 14:30 33, 1994 17:196 213, 1997 17. Keshaviah PR: The solute removal index A unified basis for 30. Van der Voort JH, Harvey EA, Braj B, Geary DF: Can the comparing disparate therapies. Perit Dial Int 15:101 104, 1995 DOQI guidelines be met by peritoneal dialysis alone in pediatric 18. Keshaviah PR, Star RA: A new approach to dialysis quantifica- patients? Pediatr Nephrol 14:717 719, 2000 tion: An adequacy index based on solute removal. Semin Dial 7: 31. Kabanda A, Goffin E, Bernard A, et al: Factors influencing 85 90, 1994 serum levels and peritoneal clearances of low molecular weight 19. Canaud B, Garred LJ, Argiles A, et al: Creatinine kinetic modelling: A simple and reliable tool for the assessment of protein proteins in continuos ambulatory peritoneal dialysis. Kidney Int 48:1946 1952, 1995