Peritoneal dialysis in children under two years of age

Nephrol Dial Transplant (2008) 23: 1747 1753 doi: 10.1093/ndt/gfn035 Advanced Access publication 28 February 2008 Short Communication Peritoneal dialysis in children under two years of age Hanne Laakkonen 1, Tuula Hölttä 1, Tuula Lönnqvist 2, Christer Holmberg 1 and Kai Rönnholm 1 1 Department of Paediatric Nephrology and Transplantation and 2 Department of Paediatric Neurology, Hospital for Children and Adolescents, University of Helsinki, Helsinki, Finland Abstract Background. Although results of peritoneal dialysis (PD) in small children have improved during recent years, the youngest children have poorer growth, more infections and higher mortality than do older children. Methods. In this retrospective study, we analysed patient records of all children under age 2 treated with continuous peritoneal dialysis (CPD) between 1995 and 2000 in Finland. Diagnoses leading to renal failure in these 23 children were congenital nephrotic syndrome of the Finnish type (13), polycystic kidney disease (4), a urethral valve (3), renal insufficiency due to neonatal asphyxia (2) and Prune-Belly syndrome (1). Of these 23, 17 (74%) were anuric. Results. The mean age at the onset of PD was 0.4 years and the mean time on dialysis 1.4 years. Hernias were diagnosed in 57%. The peritonitis rate was 1:14.5 patient-months, and 30% were peritonitis-free. Hypertension was common, and 70% had at least one period on antihypertensive medication. None of the patients had pulmonary oedema or dialysis-related seizures. The mean height standard deviation score (hsds) at the start of PD (n = 16) was 2.0 and after 9 months 1.6. Catch-up growth was documented in 64% of the patients during dialysis. Hospitalization time was 124 days/patient-year. Two patients (9%) died. Conclusions. Our results are reassuring. Mortality was low, laboratory parameters were acceptable and growth was good. Peritonitis rate was comparable to that in older children. Correction of inguinal hernia should be routinely performed; high blood pressure is still a problem. Keywords: children; complications; infant; outcome; peritoneal dialysis [2], and subsequently younger and younger children have been accepted for peritoneal dialysis (PD). Mortality, infections and poor growth are the major problems in infants compared to those in older children and adults [3 7]. A recent study by Shroff et al. showed the mortality rate in children needing renal replacement therapy to be 2.7-fold in those under 5 years compared to those older. Non-renal comorbidity was the most important mortality factor [8]. Finland has more infants on PD than do many other countries mainly because of the high incidence (1:8000 newborns) of congenital nephrotic syndrome of the Finnish type (NPHS1, CNF). The incidence of renal replacement therapy (RRT) in children under age 4 was 15.5 per million age-related population in Finland during 1995 2000, compared to 6.7 in 12 ERA-EDTA registry countries in Europe [9]. Hölttä et al. reported our experience with CPD in 34 children under age 5 between 1986 and 1994. Growth was good: the mean height standard deviation score (hsds) at the onset of PD was 2.1 and after 6 months 1.7, and mortality was only 6% (two patients) [2]. Ledermann et al. and Wood et al. [10,11] subsequently reported 20 children under 1 year and 103 children under 2 on CPD. Growth was good, with non-renal comorbidity the most important mortality-risk factor. Younger and younger children are included in PD programs, and their optimal therapy is very demanding. In this study we report complications, medication, laboratory parameters, growth, hospital stay and comorbidity in children treated with CPD before age 2. Patients and methods Introduction Continuous peritoneal dialysis (CPD) was introduced in children in the 1970s [1]. In Finland it began in the 1980s Correspondence and offprint requests to: Hanne Laakkonen, Department of Paediatric Nephrology and Transplantation, Hospital for Children and Adolescents, University of Helsinki, Helsinki, Finland. email: hanne.laakkonen@hus.fi Patients We report the outcome of 23 children (9 girls and 14 boys) under 2 years of age at the onset of CPD, treated between 1995 and 2000 in Finland. Eight of these patients were included in the study of Hölttä et al. [12]. Diagnoses leading to renal failure were congenital nephrotic syndrome of the Finnish type (NPHS1) (13), polycystic kidney disease (4), a urethral valve (3), renal insufficiency due to neonatal asphyxia (2) and Prune-Belly syndrome (1). Bilateral nephrectomy was performed in 17 children (74%), C The Author [2008]. The online version of this article has been published under an open access model. Users are entitled to use, reproduce, disseminate, or display the open access version of this article for non-commercial purposes provided that: the original authorship is properly and fully attributed; the Journal and Oxford University Press are attributed as the original place of publication with the correct citation details given; if an article is subsequently reproduced or disseminated not in its entirety but only in part or as a derivative work this must be clearly indicated. For commercial re-use, please contact journals.permissions@oxfordjournals.org

1748 H. Laakkonen et al. Table 1. Clinical details of 23 patients (mean age at the onset of PD 0.4 years) on chronic PD (mean duration 1.4 years) during 1995 2000 No. Sex Diagnosis Nephrectomy PD onset PD duration/ Peritonitis Catheter PD ended in unilat/bilat age/years years number complications 1 F CNF Bilateral 0.6 3.6 5 Yes Tx 2 M CNF Bilateral 0.5 1.0 1 No Tx 3 F CNF Bilateral 0.7 2.0 3 Yes Tx 4 M CNF Bilateral 0.5 0.9 1 No Tx 5 M CNF Bilateral 0.3 0.5 0 No Tx 6 F CNF Bilateral 0.6 0.6 1 No Tx 7 M CNF Bilateral 0.6 0.7 1 No Tx 8 F CNF Bilateral 0.6 2.7 1 Yes HD 9 M CNF Bilateral 0.7 2.1 0 Yes HD 10 M CNF Bilateral 0.5 0.7 0 No Death a 11 M CNF + MDA Bilateral 0.7 1.6 1 Yes Tx 12 M CNF + MDA Bilateral 0.6 1.1 0 Yes Tx 13 F CNF + MDA Bilateral 0.8 2.2 2 No Tx 14 M ARPKD Bilateral 0.01 0.7 2 Yes Tx 15 F ARPKD Bilateral 0.01 1.3 1 Yes Tx 16 M ARPKD Bilateral 0.01 1.9 1 Yes Tx 17 M Urethral valve 0.2 0.7 0 Yes Tx 18 M Urethral valve 0.4 1.4 1 Yes Tx 19 M Urethral valve Bilateral 0.03 2.1 1 Yes HD 20 M Prune-Belly sdr Unilateral 0.05 1.1 2 No Tx 21 F Asphyctic kidney failure 0.03 0.9 0 Yes Death b 22 F Asphyctic kidney failure 0.02 1.4 2 No Tx 23 F Polycystic kidneys 0.1 0.2 0 Yes Recovery CNF, congenital nephrotic syndrome of the Finnish type; MDA, muscular dystonia and athetosis; ARPKD, autosomal recessive polycystic kidney disease; Tx, renal transplantation; HD, haemodialysis. a Due to intracerebral haemorrhagia. b Due to congenital heart disease (CoA). unilateral in 1 (Table 1). All children were treated by the same team according to the same guidelines [12]. Methods Retrospective data came from the patient records, concerning diagnoses, operations, dialysis modality, course and complications, medication, laboratory parameters and growth. Patient 23, only 0.2 years on PD, was excluded from the growth analysis. Dialysis The criterion for starting dialysis in NPHS1 patients was a weight of at least 7 kg, at which time point the children were nephrectomized to halt their severe proteinuria. Criteria for considering a patient under 2 years of age sufficiently uraemic to start PD were a blood creatinine >300 µmol/l or failure to thrive with conservative therapy. A Tenckhoff curled 1-cuff catheter with the exit site laterally and the head of the catheter pointing up was used, and after its placement, it was flushed until the dialysate was clear. Before nephrectomy and start of daily dialysis, a recovery time of 2 weeks was, if possible, allowed after catheter application. Peritoneal dialysis was instituted during 5 7 days with increasing volumes up to 800 1000 ml/m 2 of body surface area (BSA) per exchange. Cycler machines used were PAC Xtra and Home Choice (Baxter Healthcare, FL, USA), PD100 (Gambro, Lund, Sweden) and PD Night (Fresenius Medical Care, Bad Homburg, Germany). Three patients were permanently shifted to haemodialysis (HD) and were excluded from the study after start of HD. Diet and medication Each patient had a nasogastric tube to guarantee sufficient nutrition if failing to ingest the required amount. Traditionally, gastrostomies have not been much used in our centre. The diet was infant milk and cereal formulae supplemented with glucose polymers or a combined fat and carbohydrate product and casein-based protein. Mean energy provided was 110 120% of the recommended dietary allowance, the protein intake was 2.0 3.0 g/kg/day and most was provided through the nasogastric tube. All patients received standard medication for their endstage renal failure. Calcium carbonate served as a phosphate binder and alphacalcidol as vitamin D substitution aiming at maintaining the intact parathyroid hormone (ipth) concentration lower than twice the upper limit of its normal reference range. Patients received subcutaneous erythropoietin and peroral iron supplementation to maintain adequate haemoglobin levels. Vitamin C, multivitamin formulae containing vitamin B and folic acid, and sodium supplementation were administered if clinically indicated [12]. No patient was on growth hormone therapy. The daily or weekly doses of their medication are reported after 6 months on PD. Laboratory parameters All patients visited our clinic at least every third month. We report the patients essential laboratory tests (blood haemoglobin, urea, creatinine, albumin) at 3, 6 and 9 months on PD. Adequacy of PD was measured in 20 patients with a modified 24-h dialysate and urine collection

Peritoneal dialysis in children under two years of age 1749 started at noon. The patient s night dialysis was then started some hours earlier than normally and an 8-h dwell (1000 ml/m 2, 2.27% glucose solution) preceding the peritoneal equilibration test (PET) was included in the collection [12]. We report urea Kt/V and creatinine clearance after 6 12 months in 15 patients. Statistics Data were analysed with SPSS 10.0 for Windows software (SPSS Inc., Chicago, IL, USA) and are given as mean values, except for parathyroid hormone levels, presented as medians. Statistical analyses for growth (hsds) were performed with the Wilcoxon signed-ranks test. Statistical significance was defined as P < 0.05. Results Dialysis The mean age was 0.4 years (0 0.8) at the start and 1.8 years (0.3 4.2) at the end of PD. Patients were on nightly automated PD. Anuric patients had two additional afternoon exchanges to maintain normal volaemia and blood pressure [12]. Modalities used were continuous ambulatory peritoneal dialysis (CAPD) in the beginning, when fill volume was <150 ml, and thereafter continuous cycling peritoneal dialysis (CCPD) and tidal peritoneal dialysis (TPD). With lower volumes, in CCPD and TPD, there was a low flow in the tubing causing frequent alarms and a proportionally high recirculation volume. CCPD was the first choice, but patients treated during 1995 1998 were switched to TPD for 6 months because of an ongoing study [13]. Later, some patients were switched to TPD because of outflow pain or to achieve better purification. Mean time on dialysis was 1.4 years (0.2 3.6), adding up to 377 peritoneal dialysis months. Of these months, the patients were on CAPD for 45.5 (12%), on CCPD for 250 (66%) and on TPD for 81.5 months (22%). One patient was on CAPD during the entire 11 months of dialysis. Six patients needed a catheter exchange once, three patients twice and one patient three times. Underlying causes were peritonitis or tunnel infection in seven, leakage in three, and malposition and obstruction of the catheter in one patient each. Catheter repositions were performed once in four and twice in two patients due to catheter migration (Table 1, Figure 1). Five patients were temporarily transferred (<1.5 months) to HD because of peritonitis, catheter infection, leakage or ultrafiltration (UF) failure. Three patients were shifted permanently to HD: two due to peritonitis and one due to UF failure. Peritonitis and non-infectious complications During PD, 13 patients (57%) underwent operations because of inguinal or umbilical hernias or hydrocele testis. Four of these needed two separate operations. Three patients were each operated on for an inguinal hernia before PD onset; one underwent a second operation during PD. Fig. 1. Hernias, peritonitis, tunnel and exit-site infections, catheter exchanges and repositions in 23 patients under age 2 at PD initiation. Sixteen patients suffered 26 episodes of peritonitis (Tables 1 and 2), giving a peritonitis incidence of 1 episode/ 14.5 patient-months and 0.83 episodes/patient-year. Peritonitis commenced at home in 63% and at hospital in 37%. Two episodes (7%) occurred within 2 weeks after catheter insertion. Four patients had altogether 11 tunnel infections, and three patients 7 exit-site infections (Figure 1). None of the tunnel infections resulted in peritonitis. Gram-positive and Gram-negative bacteria each accounted for 37% of dialysate cultures, and 22% were culture-negative. The four most common bacteria were Staphylococcus epidermidis (19%), Pseudomonas (15%),Staphylococcus aureus (11%) and Klebsiella (11%). Only one peritonitis (4%) was fungal. Most tunnel and exit-site infections were caused by S. aureus (44%). Exit-site infections were treated with peroral, and tunnel infections with intravenous cephalosporins. We used no prophylactic treatment for nasal carriers of S. aureus. During PD, 16 patients (70%) had at least one period on antihypertensive medication. At PD onset, 39% needed medication (11% of these with more than one medicine); at 3 months, 43% (10% with more than one medicine); at 6 months, 50% (27% with more than one medicine) and at 9 months, 53% (33% with more than one medicine) (Table 2). Thus, the proportion needing medication for hypertension and the number of preparations used increased with time. Antihypertensive treatment was a calcium channel blocker in most patients (nifedipine or felodipine), but also labetalol, propranolol, metoprolol and enalapril. Diet and medication Nutrition was provided orally but intake was minimal. Thus, close to 100% was received by nasogastric tube. After renal transplantation, normal peroral intake was achieved in all patients within a few weeks. Medication was adjusted every 3 weeks and is reported after 6 months on PD (n = 22). The average daily dose of calcium (as calcium carbonate) was 69 mg/kg (8 189). Alphacalcidol was mostly administered as pulse therapy at an average dose of 2.0 µg/week/patient (0.2 5.4) or 0.23 µg/kg/week (0.03 0.55). The mean erythropoietin dose was 283 IU/kg/week (68 577) and mean oral iron (Fe 2+ ) dose 3.6 mg/kg/day (0.8 6.0). Sodium chloride

1750 H. Laakkonen et al. Table 2. PD patients under 5 and 2 years of age at onset, treated in Finland; comparison of two previous studies (Hölttä et al. [2,12]) with present data. Age, PD duration, height SDS and hospitalization time given as mean values Hölttä et al. 1997 [2] Hölttä et al. 2000 [12] Present study <5 years <5 years <2 years Number of patients 34 10 a 23 a Age at onset (years) 1.6 ± 1.0 1.0 ± 0.6 0.4 ± 0.3 PD duration (years) 0.78 0.8 1.4 Height SDS after 6 months on PD 1.7 ± 1.5 1.1 ± 1.1 1.4 ± 1.7 Hospitalization (days/patient-year) 150 95 124/93 b Antihypertensive medication 64% 30% 70% Pulmonary oedema 41% 0% 0% Seizures 26% 0% 0% Peritonitis frequency 1/7.3 months 1/9.4 months 1/14.5 months Catheter exchange (% of patients) 21% 10% 43% Hernias (% of patients) 29% Not reported 57% Mortality 6% 10% 9% a Eight patients are included in both studies. b Excluding three patients spending their whole PD time in hospital: one because of severe heart disease and two for social reasons. Table 3. Laboratory parameters in 22 children under 2 years of age at the start of CPD (mean ± 1 SD) Time on CPD Laboratory parameters 3 months (n = 22) 6 months (n = 21) 9 months (n = 16) Haemoglobin (g/l) 113 ± 13 109 ± 16 114 ± 11 Creatinine (µmol/l) 353 ± 133 397 ± 154 414 ± 179 BUN (mmol/l) 41 ± 13 39 ± 11 37 ± 10 Sodium (mmol/l) 140 ± 5 139 ± 3 139 ± 3 Potassium (mmol/l) 4.9 ± 0.6 4.8 ± 0.8 4.6 ± 0.8 Bicarbonate (mmol/l) 25 ± 4 26 ± 4 28 ± 4 Calcium ion (mmol/l) 1.31 ± 0.11 1.25 ± 0.08 1.29 ± 0.06 Phosphate (mmol/l) 1.93 ± 0.54 1.66 ± 0.61 1.31 ± 0.42 Alkaline phosphatase (U/L) 1201 ± 328 1332 ± 401 1421 ± 546 Prealbumin (mg/l) 419 ± 96 393 ± 157 429 ± 72 Albumin (g/l) 31 ± 6 29 ± 7 33 ± 5 Triglycerides (mmol/l) 3.2 ± 1.2 3.7 ± 1.5 2.8 ± 1.4 Cholesterol total (mmol/l) 5.6 ± 1.6 5.6 ± 1.2 5.7 ± 1.5 PTH intact (ng/l), median (range) 122 (5 521) 287 (5 1124) 112 (14 978) concentrate was used in 16 patients to compensate for intraperitoneal losses of sodium at an average dose of 3.1 mmol/kg/day (0.7 8.2). Their dose was adjusted to sodium blood levels to maintain the sodium level >132 mmol/l. Laboratory parameters Haemoglobin, albumin and calcium concentrations during dialysis were acceptable (Table 3). In the beginning of PD, the ipth levels were less than twice the upper limit of the reference range in 95% of the patients (n = 20), after 3 months in 57% (n = 21) and after 6 months in 19% (n = 21). The mean adequacy parameters at 9 months on PD (6 12 months) were urea Kt/V 3.2 ± 1.0 and creatinine clearance 66.8 ± 23.3 L/week/1.73 m 2. Anuric patients had aureakt/v3.0± 0.5 and creatinine clearance of 62.2 ± 17.2 L/week/1.73 m 2, and the patients with residual renal function had 4.0 ± 2.2 and 85.0 ± 39.2 L/week/1.73 m 2. Growth The hsds for each patient are given in Figure 2. The child with heart insufficiency (operated aortic coarctation and multiple VSD, No. 21 in Table 1) had poor growth with an hsds of 6.5 at 6 and 9 months. In patients who had been at least 9 months on PD (n = 16), the mean hsds at onset of PD was 1.9 ± 1.2 and after 9 months 1.6 ± 1.8. Excluding one patient with a congenital heart defect, the mean hsdsimprovedto 1.8 ± 1.1 and 1.3 ± 1.2 at PD onset and 9 months, respectively. Comparing the mean hsds in the beginning and at the end of PD (after 0.5 to 3.6 years) in 22 patients, the mean hsds were 1.5 ± 1.4 and 1.2 ± 1.7. Excluding the patient with heart failure, the mean hsds were 1.3 ± 1.3 and 0.9 ± 1.3. These differences are, however, not statistically significant. Catch-up growth during dialysis was documented in 14 patients (64%), and only one of these had residual renal function (RRF). No statistical difference appeared in albumin, PTH, calcium or phosphate levels between patients showing catch-up growth or poorer growth. The five children with RRF grew no better than did anuric children. Their mean hsds in the beginning of PD was 1.5 ± 1.6 ( 1.0 ± 1.1 without patient 21) and at the end 2.2 ± 2.8 ( 1.3 ± 1.6), compared to 1.4 ± 1.3 and 0.8 ± 1.2 in 17 anuric children. The change in hsds during whole dialysis time in children with RRF was 0.7 ± 1.5 ( 0.2 ± 1.2 without patient No. 21) and

Peritoneal dialysis in children under two years of age 1751 Fig. 2. Growth (height SDS) in 22 children under 2 years of age, 6 and 3 months before, at onset and during PD; each line represents an individual patient. Star marks the patient with congenital heart disease. 0.6 ± 1.4 in anuric children.mean body mass index (BMI) was 16.7 ± 2.5 in the beginning and 17.8 ± 2.3 after 6 and 17.7 ± 2.0 after 9 months on PD (n = 16). Hospitalization and non-renal abnormalities Hospitalization time was 124 days/patient-year. Three patients had to spend their entire dialysis period in hospital: one because of severe congenital heart disease and two for social reasons. Excluding these three, the hospitalization time fell to 93 days/patient-year. Treatment of peritonitis and exit-site infections required 11 hospital days, other complications (such as hypervolaemia) 11 days and social reasons 15 days/patient-year. Excluding those three patients living in hospital, the proportion of social reasons fell to 7 days/patient-year. Mortality was 9% (two patients) during the 377 patientmonths, which means the mortality of one patient every 16 PD years. One patient with normal blood pressure developed a subarachnoidal haemorrhage, and the second died of congenital heart disease with severe heart insufficiency (equal to comorbidity). Seven additional patients had comorbidity: three had, in addition to CNF, muscular dystonia with athetosis (MDA), two developmental delay, one periportal fibrosis and multiple pancreatitis, and one patient haemiplegia, periportal fibrosis and Asperger syndrome. Two CNF patients with associated MDA have died later, after renal transplantation [14]. One patient suffered developmental delay and visual disability because of hypo- and hypervolaemic periods during dialysis. Discussion Nowadays, we can treat even newborns with PD prior to renal transplantation, but these make up the most challenging patient group. Problems are frequent infections, poor growth [7] and a higher mortality rate [8] than for older children. In our study of 23 children whose CPD was started during the first 2 years of life, between 1995 and 2000, 57% needed hernia operations during PD, three even before PD initiation (Table 2). Hernias and catheter-related problems are common during PD, but few reports exist as to hernias in infants. Jander et al. [15] reported 29 infants on PD in Poland: 31% needed a hernia operation during their study period. Hölttä et al. [2] documented hernias in 29% of 34 children under age 5 treated between 1986 and 1994. Consequently, it seems that these infants are more susceptible to hernia, and in the youngest patients one ought to consider any need for a hernia operation at catheter insertion. Catheter-related technical problems are also common. Rees et al. [16] reported that of their 20 infants, 12 (60%) needed at least one catheter exchange. Other studies report a revision ratio (number of revisions/number of accesses) of 0.2 (20%) in children and adolescents [7,17]. Of our patients, 10 (43%) needed a catheter exchange at least once (revision ratio = 0.65). The NAPRTCS report for 2001 shows a peritonitis rate of 1 episode/11.8 patient-months in children under 1 year of age, and 1:14.7 for all children [5]. Boehm et al. [18] reported a peritonitis rate of 1:9.8 in children under 2 and 1:23.1 in older children. Their study also showed a peritonitis rate of 1:8.5 in anuric children compared to 1:19.7 in children with residual renal function, demonstrating that anuria is a risk factor for peritonitis. Anuric children typically need day-exchanges and in consequence are exposed to more tubing connections, a fact which may in part explain their higher peritonitis ratio. Between 1986 and 1994 we saw one episode of peritonitis every 7.3 patient-months in children under 5 years [2] compared to the present study rate of only 1 episode/14.5 patient-months, with 30% remaining peritonitis-free. Our anuric patients had a peritonitis ratio of 1:14.8, and those with residual renal function 1:14.1. Of our patients, 74% were anuric, but they had a peritonitis rate comparable to those in older, non-anuric children. Gram-positive bacteria are the most common cause of peritonitis [2,7,12,19], but Gram-negative bacteria were equally common in our patients. The rate of culturenegative peritonitis was 22%, which is comparable to the rate of 21.6% in NAPRTCS report in 2000 [19]. Most infections occurred at home, but 35% even had their first episode within 1 month of PD start in hospital. For infants on PD, few reports deal with blood pressure and its medication. Hypervolaemia and related complications during dialysis are important predictors for neurodevelopmental outcome after renal transplantation [20]. During 1986 1994, 69% of our children under age 5 on PD were on antihypertensive medication [2]. In the later study of Hölttä et al. the percentage was only 30% in the same age group [12]. In the present study with younger children, 70% of the patients were on antihypertensive medication at some point in dialysis. Thus, medication was more aggressive, but no patient developed pulmonary oedema or dialysis-related seizures except one with hypertensive and hypovolaemic periods with later developmental delay and visual impairment (patient 1). In small children, estimation

1752 H. Laakkonen et al. of volaemia is difficult, and we need more accurate methods for its evaluation. Meanwhile, it is presumably safer to keep volaemia on the higher side in infants, even at the cost of more antihypertensive medication. One could also speculate that those patients who are high transporters are more prone to hypervolaemia. This issue is addressed in a prospective study in progress. The creatinine clearances reported here are an overestimation as the 8-h dwell prior to the peritoneal equilibration test was included in the calculations. We are going to report peritoneal characteristics, prescriptions and clearances in infants in detail in a subsequent report in preparation. Growth was good, and all children had nasogastric-tube feeding to ensure sufficient energy and protein and vitamin intake, as well as adequate medication. Some centres successfully use gastrostomy for adequate feeding [21]. Schroff et al. [8] report growth deficit to be most severe in the youngest patients (under 5). At the beginning of PD, mean hsds was 3.6 in their patients and final hsds 2.6. In our patients, these values were 1.5 and 1.2. Catch-up growth was documented in most patients without growth hormone treatment. Calcium-phosphate balance has to be maintained, and in our experience intermittent vitamin D seems better for growth, and a PTH concentration two to four times normal to be acceptable [22]. Chadha et al. in 2001 reported 24 PD patients (ages 7 days to 16 years) with a wide variety of diagnoses but similar adequacy parameters. Only nine of these showed catch-up growth, seven of whom had residual renal function, considered the main reason for their better growth [23]. However, of our 22 patients under age 2 at start of dialysis, 5 had residual renal function but not superior growth. As the mean waiting time for renal transplantation, at our institution, is only 9 months, in nine patients (39%) the follow-up time was <1 year. In addition, 16 of our patients had hsds > 2 at the start of PD. Taking these limitations into account, it seems clear that one can achieve normal growth in infants on PD with early institution of PD, appropriate medication and adequate nutrition. We thus emphasize the importance of early initiation of PD in uraemic infants. For these youngest patients hospitalization time is still long, although it has decreased recently. More frequent catheter exchange and hernia surgery as well as comorbidity and social causes increase hospitalization in the youngest age group, but in the long run it is important to optimize treatment during infancy to guarantee subsequent normal growth and development later, even at the cost of more frequent hospital visits. Our mortality was 9%. Schroff et al. [8] reported a mortality rate of 10% during dialysis and 17% during their whole follow-up period, giving a relative risk for death 2.7 times higher than that in their older children. Non-renal comorbidity is a significant risk factor for higher mortality [8,10,11]. Both children who died during PD at our centre had comorbidities, as did the two patients who died later, after renal transplantation. In conclusion, even infants can today be successfully treated by CPD, but the long-term effects are still unknown. We thus started a prospective study in infants on PD in 2001, with strict control of PD, detailed neurological followup and quality-of-life assessment of the whole family, including the parent child relationship; today, home visits by a dialysis nurse are routine. The overall results of the present study, however, have proven reassuring, concerning survival, growth and metabolic control. Conflict of interest statement. None declared. References 1. Oreopoulos DG, Katirtzoglou A, Arbus G et al. Dialysis and transplantation in young children. Br Med J 1979; 1: 1628 1629 2. Hölttä TM, Rönnholm KA, Jalanko H et al. Peritoneal dialysis in children under 5 years of age. Perit Dial Int 1997; 17: 573 580 3. Bunchman TE. Chronic dialysis in the infant less than 1 year of age. Pediatr Nephrol 1995; 9(Suppl): S18 S22 4. Verrina E, Edefonti A, Gianoglio B et al. A multicenter experience on patient and technique survival in children on chronic dialysis. Pediatr Nephrol 2004; 19: 82 90 5. Neu AM, Ho PL, McDonald RA et al. Chronic dialysis in children and adolescents: the 2001 NAPRTCS annual report. Pediatr Nephrol 2002; 17: 656 663 6. Howard RL, Millspaugh J, Teitelbaum I. Adult and pediatric peritonitis rates in a home dialysis program: comparison of continuous ambulatory and continuous cycling peritoneal dialysis. Am J Kidney Dis 1990; 16: 469 472 7. Warady BA, Hebert D, Sullivan EK et al. Renal transplantation, chronic dialysis, and chronic renal insufficiency in children and adolescents: the 1995 annual report of the North American Pediatric Renal Transplant Cooperative Study. Pediatr Nephrol 1997; 11: 49 64 8. Shroff R, Rees L, Trompeter R et al. Long-term outcome of chronic dialysis in children. Pediatr Nephrol 2006; 21: 257 264 9. Van Der Heijden BJ, van Dijk PC, Verrier-Jones K et al. Renal replacement therapy in children: data from 12 registries in Europe. Pediatr Nephrol 2004; 19: 213 221 10. Ledermann SE, Scanes ME, Fernando ON et al. Long-term outcome of peritoneal dialysis in infants. J Pediatr 2000; 136: 24 29 11. Wood EG, Hand M, Briscoe DM et al. Risk factors for mortality in infants and young children on dialysis. AmJKidneyDis2001; 37: 573 579 12. Hölttä T, Rönnholm K, Jalanko H et al. Clinical outcome of pediatric patients on peritoneal dialysis under adequacy control. Pediatr Nephrol 2000; 14: 889 897 13. Hölttä T,Rönnholm K, Holmberg C. Adequacy of dialysis with tidal and continuous cycling peritoneal dialysis in children. Nephrol Dial Transplant 2000; 15: 1438 1442 14. Laakkonen H, Lönnqvist T, Uusimaa J et al. Muscular dystonia and athetosis in six patients with congenital nephrotic syndrome of the Finnish type (NPHS1). Pediatr Nephrol 2006; 21: 182 189 15. Jander A, Nowicki M, Tkaczyk M et al. Chronic peritoneal dialysis in infants preliminary results of the multicenter survey. Przegl Lek 2006; 63(Suppl 3): 72 74 16. Rees L. Management of the infant with end-stage renal failure. Nephrol Dial Transplant 2002; 17: 1564 1567 17. Lerner GR, Warady BA, Sullivan EK et al. Chronic dialysis in children and adolescents: the 1996 annual report of the North American Pediatric Renal Transplant Cooperative Study. Pediatr Nephrol 1999; 13: 404 417 18. Boehm M, Vecsei A, Aufricht C et al. Risk factors for peritonitis in pediatric peritoneal dialysis: a single-center study. Pediatr Nephrol 2005; 20: 1478 1483 19. Furth SL, Donaldson LA, Sullivan EK et al. North American Pediatric Renal Transplant Cooperative Study. Peritoneal dialysis catheter infections and peritonitis in children: a report of the North American Pediatric Renal Transplant Cooperative Study. Pediatr Nephrol 2000; 15: 179 182

Peritoneal dialysis in children under two years of age 1753 20. Qvist E, Pihko H, Fagerudd P et al. Neurodevelopmental outcome in high-risk patients after renal transplantation in early childhood. Pediatr Transplant 2002; 6: 53 62 21. Ledermann SE, Spitz L, Moloney J et al. Gastrostomy feeding in infants and children on peritoneal dialysis. Pediatr Nephrol 2002;17: 246 250 22. Saarinen T, Arikoski P, Holmberg C et al. Intermittent or daily administration of alpha-1-calcidol for nephrectomised infants on peritoneal dialysis? Pediatr Nephrol 2007; 22: 1931 1938 23. Chadha V, Blowey DL, Warady BA. Is growth a valid outcome measure of dialysis clearance in children undergoing peritoneal dialysis? PeritDialInt2001; 21(Suppl 3): S179 S184 Received for publication: 14.2.07 Accepted in revised form: 18.1.08