Article Text

Download PDFPDF


Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.


S. D. Marks, R. Davies, M. J. Dillon, P. Duffy, I. Gordon, G. Hamilton, R. HLord, C. McLaren, D. Roebuck, V. Shah, R. Shroff, S. E. Stephen, K. Tullus.Great Ormond Street Hospital for Children NHS Trust, London, UK

Aims: To evaluate the clinical outcomes, morbidity, and mortality of children undergoing medical and surgical interventions in renovascular hypertension in a tertiary and quaternary paediatric nephrology service over the past 30 years.

Methods: Patients were identified through searches of the Paediatric Nephrology database at Great Ormond Street Hospital for Children. Exclusion criteria included hypertension secondary to vasculitis (such as polyarteritis nodosa) and transplant renal artery stenosis.

Results: Full data were available for a total of 109 patients (70 male (64%)) who presented at aged 0.5–17.0 (median 7.7) years with renovascular hypertension. Twenty three (21%) patients were asymptomatic with hypertension noted as an incidental finding and 27 (25%) diagnosed secondary to a hypertension associated syndrome (namely neurofibromatosis type 1, Williams, and velo-cardial facial syndromes). Systolic blood pressure at presentation was 120–270 (median 167) mm Hg. Patients required between two and nine (median 4) antihypertensive agents. There was evidence of end organ damage with 73 (67%) patients with left ventricular hypertrophy, 36 (33%) with hypertensive retinopathy, and 40 (37%) with proteinuria. The estimated glomerular filtration rate was 67–122 (median 108) mls/min/1.73 m2. Medical management was undertaken in all patients by consultant paediatric nephrologists. Forty six (42%) children underwent 73 interventional radiological and/or surgical procedures for treatment of hypertension, including transluminal angioplasty (in 33), nephrectomy (in 17), and revascularisation and autotransplantation procedures (in 23) by interventional radiologist, urologists, and transplant and vascular surgeons. Twenty three (50%) patients were deemed to be cured as they were normotensive off all antihypertensives after therapeutic interventions. This was possible in 47% of nephrectomies, 27% of angioplasties, and 26% of surgical procedures. The major complications of angioplasty included one fatal haemorrhage from graft tear, stent migration, thrombosis, haematoma, and focal defects on DMSA scan. There were no major complications post-nephrectomy but one patient had a thrombosis of the affected kidney after a revascularisation procedure.

Conclusions: 10% of childhood hypertension is due to renovascular disease. While many patients present with uncontrollable hypertension, the possibility of curative therapeutic interventions with angioplasty and surgery remain on an individual patient basis with a multidisciplinary team decision. Out of 46 children who underwent interventional procedures, 50% were cured off treatment at follow up.


R. Lennon1, G. I. Welsh3, I. Hers3, M. Sabin2, J. Shield2, P. W. Mathieson1, M. A. Saleem1.1Academic and Childrens Renal Unit, Bristol, UK; 2Paediatric Endocrinology, Bristol Childrens Hospital, Bristol, UK; 3Department of Biochemistry, Bristol University, Bristol, UK

Background: We have previously demonstrated that human podocytes in vitro are insulin sensitive and able to rapidly transport glucose utilising the glucose transporter GLUT 4. We hypothesised that, as with other insulin responsive cells, it would be possible to induce insulin resistance in human podocytes. Insulin resistance is a feature of type 2 diabetes mellitus, and is strongly associated with diabetic nephropathy. Conditionally immortalised human podocytes were grown to confluency and fully differentiated. Cells were incubated for 24 hours in serum free medium containing 5% BSA (fatty acid free) and 750 μM palmitate (a saturated free fatty acid). For SDS-PAGE and immunofluorescence, cells were stimulated with insulin (200 nM) and glucose (25 mM). Cells for glucose transport were stimulated with insulin alone before adding labelled glucose. The PPARγ agonist ciglitazone (20 μM) was added to a proportion of the cultures.

Glucose Transport: The ratio of insulin stimulated to basal glucose uptake was a mean of 1.9 for control cells and this effect was significantly reduced by treatment with palmitate for 24 hours (p 0.003). There was less of a reduction in cells treated with palmitate for less than 12 hours and there was no reduction in glucose uptake in cells treated with BSA alone. With ciglitazone, glucose uptake was increased in control cells without insulin stimulation suggesting an independent effect on glucose uptake. There was no improvement in glucose uptake in cells co-treated with palmitate and ciglitazone.

Western Blot: Using the antiphosphotyrosine antibody 4G10, there was a clear insulin effect in control cells, including upregulation of a band at the level of the insulin receptor substrate 1 (IRS1). These changes were not present in palmitate treated cell lysates.

Immunofluorescence: Control cells demonstrated cytoplasmic GLUT 4 staining compared with stimulated cells that had a marked increase in cell membrane GLUT 4. There was minimal change in GLUT 4 staining in palmitate treated cells following stimulation.

Conclusion: Human podocytes in vitro have impaired glucose transport in response to insulin stimulation following treatment with palmitate. Furthermore, there is a reduction in phosphorylation of proteins involved in insulin signalling and a decrease in GLUT 4 transport to the cell membrane. These findings suggest that free fatty acid induces insulin resistance in human podocytes.


P. A. Brogan1, K. Tullus2, W. Van’t Hoff2, R. Trompeter2, L. Rees2.1Institute of Child Health, London, UK; 2Great Ormond St Hospital, London, UK

Aims: To describe the incidence, cause, and complications of pancreatitis in a single centre population of children with CRF.

Methods: Local and hospital-wide database searches and retrospective case note review.

Results: From April 1995 to January 2004, 22/1010 children with CRF (GFR <80 mls/min/1.73 m2) developed acute pancreatitis, representing a prevalence of 2.1%. Detailed data were available from 17 children, with 18 episodes of pancreatitis. There were 13 females and 4 males of median age 11.8 years (1.0–18.2 years). Median (range) pre-morbid calculated GFR was 54 (<10–80) mls/min/1.73 m2; 8/17 had a GFR less than 15 mls/min/1.73 m2. The cause of CRF was congenital malformation of the renal tract (n = 5); haemolytic uraemic syndrome (HUS (n = 3)); systemic lupus erythematosus (n = 2); congenital nephrotic syndrome (n = 2); metabolic disease (n = 2); focal segmental glomerulosclerosis (n = 1); and presumed calcineurin toxicity (n = 2; 1 cardiac transplant; 1 bone marrow transplant). 5/17 were undergoing dialysis 3 PD (prevalence of pancreatitis 2.7%); 2 HD (pancreatitis prevalence 2.2%). Three children developed pancreatitis a median of 23 months post renal transplantation, representing a pancreatitis prevalence of 0.9% in this transplant population. Acute on CRF during the episode of pancreatitis was common, with 5/12 children not previously on dialysis needing temporary dialysis (4 PD, 1 HD). The diagnostic sensitivity of pancreatic enzymes for the identification of pancreatitis within 48 hours of onset of symptoms was 89% for hyperamylasaemia alone, rising to 100% when hyerlipasaemia was also considered. This sensitivity fell when positive diagnosis was defined as enzyme levels three times the upper limit of normal. A major causal factor for pancreatitis was identified in 13/18 episodes: D+ HUS (n = 3); peritonitis on PD (n = 3); lupus pancreatitis (n = 2); post gastrostomy insertion (n = 1); bowel perforation (n = 1); congenital malformation of the pancreas (n = 1); chronic pancreatitis with pseudocyst (n = 1); and methyl malonic acidaemia (n = 1). Additional risks identified were hypercalcaemia (64% of cases), and poly-pharmacy. Pancreatitis was associated with considerable acute morbidity although no child required pancreatic resection. 2/17 had chronic/recurrent pancreatitis and 3/17 developed diabetes mellitus. One child died 6 months following the episode of pancreatitis.

Conclusion: There is an increased prevalence of acute pancreatitis in children with CRF. The cause of this is multifactorial. The renal prognosis following pancreatitis is generally favourable with recovery of acute on CRF, although there is an increased risk of diabetes following pancreatitis.


N. J. A. Webb, S. Stephens, H. Maxwell, M. Fitzpatrick, D. Hughes, L. Rees, N. Moghal1, K. Franks.British Association for Paediatric Nephrology, UK

Background: UK Renal Association guidelines (2002) state that fasting cholesterol and triglyceride levels should be measured in all children commencing renal replacement therapy and at annual intervals. Despite this guideline, there appears to be much variation in practice nationally.

Aims: To determine the frequency with which lipid levels are monitored in paediatric renal transplant recipients.

Methods: A single audit nurse reviewed the medical records and results databases of all patients managed by the tertiary paediatric nephrology units in Birmingham, Glasgow, Leeds, Liverpool, London (Great Ormond Street), Manchester, and Newcastle. Data were collected regarding the frequency with which plasma lipids had been measured during the 12 month period 1 April 2003 to 1 April 2004, along with other basic demographic data.

Results: A total of 308 sets of medical notes and results were reviewed. A total of 183 children (59.4%) had undergone measurement of their plasma lipids during the 12 month study period. The percentage of patients undergoing measurement of in each of the seven units varied between 19.0% and 96.9%. The median (IQR) total cholesterol level was 4.3 (3.7–5.0) mmol/l. The highest recorded level was 7.8 mmol/l. Only seven (2.3%) children were receiving statin therapy, six of these being managed in one centre.

Conclusions: There appears to be much variation in practice regarding the monitoring of lipid levels in the UK paediatric renal transplant population, despite clear national guidelines. However, lipid levels in those monitored appear to be well controlled, despite very low rates of statin use.


I. R. Witherden1, R. Viney2, L. Ni1, M. Ladomery2, P. W. Mathieson1, M. A. Saleem1.1Academic and Children’s Renal Unit, University of Bristol, Bristol, UK; 2Bristol Genomics Research Institute, University of the West of England, Bristol, UK

Background: WT1 is a transcription factor that is vital for nephrogenesis, and for reasons that remain unclear is persistently expressed in the glomerular podocyte. Its strong expression during development suggests it may play a part in podocyte differentiation.

Methods: We developed conditionally immortalised podocytes from a child with Denys Drash Syndrome (DDS), with a mutation in the zinc finger region of WT1. We compared their phenotype with wild type podocytes (WT), with the ability to differentiate when thermoswitched from 33°C to 37°C. Proliferating DDS cells at 33°C, and differentiated cells at 37°C displayed an enhanced fibroblastic morphology compared to WT. Expression of WT1 by western blotting and immunofluorescence (IF) was similar in both cell lines, and decreased in the differentiated state. Remarkably, there was a massive increase in PAX2 (transcription factor) expression in DDS cells. We also detected increased expression of the epidermal growth factor receptor, a putative WT1 target molecule, and stimulation of DDS cells with epidermal growth factor resulted in a dramatic morphological change to smooth muscle like cells.

Results: In DDS cells, consistent with the fibroblastic morphology, there was enhanced expression of myofibroblastic markers such as α smooth muscle actin, fibronectin, and myosin. Cytoskeletal patterning of DDS cells was altered, with loss of actin cytoplasmic filaments, and decrease in synaptopodin expression. Consistent with this, the normal distribution of nephrin and podocin, slit diaphragm proteins associated with actin, was disrupted, though both continued to be expressed. In vivo staining of renal cortex from the same patient showed nephrin and podocin to be expressed peripherally in the glomerulus, in areas of maximal sclerosis.

Conclusion: These observations suggest that WT1 acts in a manner to suppress mesenchymal transformation in the mature podocyte, and we propose that some areas of mesangial sclerosis typically seen in DDS biopsies may in fact represent transdifferentiated podocytes. This could represent a novel concept of the manner in which the podocyte responds to injury.


A. J. McKie1, E. Porter1, M. P. White1, J. H. McColl2, N. Aladangady3, A. Watt4, T. J. Beattie4.1Southern General Hospital, Glasgow, UK; 2University of Glasgow, Glasgow, UK; 3Homerton University Hospital, London, UK; 4Royal Hospital for Sick Children, Glasgow, UK

Aims: To examine the natural history of nephrocalcinosis and the longer term effects of prematurity and nephrocalcinosis on renal function.

Methods: Fourteen children from a cohort of ex-very low birth weight (LBW) preterm babies who had nephrocalcinosis as neonates were reviewed at the age of 5–7 years. Fourteen controls matched for sex, gestation, and birth weight were also studied. Ethics approval and informed written consent were obtained. Height, weight, and blood pressure were measured and history was taken regarding any renal symptomatology. Renal ultrasound scan was performed looking for persisting calcification, kidney size, and morphology. An early morning urine sample was obtained and routine urinalysis, osmolality, and albumin/creatinine, protein/creatinine, phosphate/creatinine, calcium/creatinine, β2-microglobulin/creatinine, and oxalate/creatinine ratio analyses were undertaken. Blood was taken from the study group for plasma creatinine, calcium, and phosphate and estimated GFR (Schwartz-Haycock formula) and TmP/GFR were calculated. Statistical analysis was carried out on a group basis using the Mann–Whitney CI.

Results: Mean age at follow up was 6.8 years (range 5.8 to 7.7). Three cases and one control had systolic hypertension and one case had systolic and diastolic hypertension. One case had previous confirmed urinary tract infection. Renal ultrasound scan was performed in 24 children (12 cases and 12 controls). Three of the 12 cases (25%) and one control had evidence of nephrocalcinosis. Unilateral impairment of kidney growth by >2 SD was seen in two cases and one control and one case and one control had bilaterally small kidneys. All patients achieved an early morning urine osmolality >700 mosm/kg. Six children had a calcium/creatinine ratio >0.7 mmol/mmol (two cases and four controls). All cases had a normal GFR with a median of 132.6 ml/min/1.73 m2 (104–173). The median TmP/GFR was 1.22 mmol/l (0.73–1.61) with two children having levels below the normal range for age. These children did not have persisting nephrocalcinosis. There were no statistically significant differences in clinical characteristics, urine biochemistry, kidney size, or symptoms of renal disease between the two groups.

Conclusions: Resolution of nephrocalcinosis occurred in 75% of cases in our series. TmP/GFR was decreased in two children indicating possible proximal renal tubular dysfunction. Normal GFR and adequate renal concentrating capacity were demonstrated suggesting normal glomerular and distal tubular function. 21% of the children had an elevated calcium/creatinine ratio although there was no significant difference between study and control groups with regard to this parameter.


N. Gittins1, M. G. Coulthard1, J. N. S. Matthews2.1Royal Victoria Infirmary, Newcastle Upon Tyne, UK; 2University of Newcastle Upon Tyne, Newcastle Upon Tyne, UK

Background: It is conventional to instil heparin into central venous haemodialysis lines as a “lock” to prevent their occlusion with blood clots. Using alteplase for clotted lines, we wondered whether this plasminogen activator may be superior to heparin as a routine lock to prevent clots.

Aims: To establish whether an alteplase or a heparin lock is more effective in preventing clot formation in central lines between haemodialysis sessions.

Methods: We studied nine children that we dialyse through central venous lines. We used either alteplase 1 mg/ml or heparin 5000 U/ml, using 1.5 or 2.0 ml, according to their line volume. Each child was their own control, and received an equal number of heparin and alteplase locks using a within-patient, multiperiod, double blind crossover design, over 10 weeks. We instilled a lock at the end of each dialysis session, and aspirated and weighed any clots that had formed at the start of the next. The randomised order allocation and statistical analysis took account of the variations in interdialytic intervals.

Results: Seven children were dialysed three times a week, and twice a weekly, one of whom was transplanted during week 9. One child required an alteplase infusion to clear a blocked line. The mean weight of clot using alteplase was only half that when using heparin (14.1 v 29.4 mg). The difference in clot weight had 95% CI of 8.1 and 21.1, and p<0.001.

Conclusion: Alteplase is a significantly more effective lock than heparin in preventing clot formation between haemodialysis sessions. As presently available it is expensive for single use, but much cheaper if it is prepared aseptically into 2 ml syringes and stored frozen until required clinically. We now use alteplase locks routinely in place of heparin, and advocate this practice. We believe this will reduce the incidence of blocked lines, and consequent alteplase infusions and line replacement surgery.


A. Waters, W. Van’t Hoff, M. Hiorns, R. J. O’Neill.Great Ormond Street Children’s Hospital, London, UK

Introduction: X-linked hypophosphataemic rickets is characterised by hypophosphataemia, renal phosphate wasting, impaired production of 1,25 dihydroxyvitamin D3, and abnormal osteoblastic function. Despite oral phosphate and 1,25 dihydroxyvitamin D3 treatment, many patients have suboptimal growth. Nephrocalcinosis is a potential complication of treatment. We assessed whether age at treatment onset affects outcome, the incidence of nephrocalcinosis, and potential risk factors in our study population.

Aims: Patients were included if they had more than 2 years’ follow up data. Of 47 rickets patients, 44 had primary inherited hypophosphataemic rickets. We excluded 13 patients with inadequate follow up (n = 8) and those that not been treated by the primary consultant (n = 5). The study group included children with positive family history who tended to be treated at an earlier age compared with those who were index cases within their family. Growth data, biochemistry and ultrasonographs of 31 patients with XLHR were analysed retrospectively. For growth analysis, patients were divided into two groups based on age at treatment onset (group 1: ⩽2 years; group 2: >2 years). Nephrocalcinosis was reported and graded according to the Patriquin and Robataille scoring system by two independent radiologists.

Methods: The median height z score was higher in group 1 (n = 17) than in group 2 (n = 14) at treatment onset (SD −1.24 v −1.85; p 0.07), at the end of the second year of treatment (−0.93 v −1.95) but similar at the time of data collection (−1.4 v −1.64; p 0.65).The change in height SDS after the first year of treatment in both groups was not statistically significant (0 v 0.11; p 0.45). The number of serum phosphates within the normal age referenced range was higher in group 1 than group 2 (median %13.3% v 0%) during the first two years of treatment but this did not have a significant impact on growth at data collection. Nephrocalcinosis was detected in 21/31 patients (67.7%) which compares with previous studies. 45% (9/20) were classified as mild, 25% (5/20) as moderate, and 30% (6/20) as severe.

Results: Results in the table below are expressed as median scores.

Conclusion: Early treatment was not associated with improved growth. But whereas untreated patients show progressive growth failure, the median height z score was maintained within 2 SD in our compliant patients (80%). Girls showed a tendency toward better growth than boys. Higher doses in the patients with no nephrocalcinosis may be due to poorer compliance.


N. J. A. Webb, D. Roberts, R. Preziosi, B. G. Keevil.Royal Manchster Children’s Hospital, Manchester, UK

Background: Regular monitoring of tacrolimus (tac) levels is an essential component of post-transplant follow up in children receiving this drug. This is usually performed at routine clinic visits, though additional monitoring may be necessary, for example after dose changes. This may be difficult where the family home is distant from the transplant centre. We have developed a tandem mass spectometry (T-MS) methodology which measures whole blood tacrolimus levels using a 10 µl sample, thus potentially allowing the use of fingerprick blood sampling. Most children generally prefer this and there is potential for samples to be collected at home and mailed to the laboratory. The aims of this study were to determine the degree of relationship between fingerprick and venous tacrolimus levels measured by T-MS and venous levels measured using the Abbott IMx Tacrolimus II assay (IMx). This has not previously been investigated.

Methods: Blood samples were collected from children on five separate occasions at the time of outpatient review. In addition to the routine venous IMx sample, a further venous and fingerprick sample were collected for tacrolimus measurement by T-MS. Samples were mailed to the central laboratory.

Results: 172 sets of triplicate samples were collected from 36 children (33 kidney, 2 kidney-pancreas, and 1 kidney-heart). All were receiving tacrolimus with prednisolone and azathioprine (21) or MMF (7). Fingerprick sampling was well tolerated; only 2 children requested withdrawal from the study because they found sampling uncomfortable. Linear regression analysis showed highly significant relationships between T-MS venous and IMx venous (r2 = 0.83), T-MS fingerprick and T-MS venous (r2 = 0.85), and T-MS fingerprick and IMx venous (r2 = 0.71) samples (all p<0.0001). The association between T-MS fingerprick and IMx venous values was stronger for values below 10 ng/ml (r2 = 0.62, p<0.0001) than for values above 10 ng/ml (r2 = 0.33, p 0.002). The slope of the T-MS fingerprick and IMx venous regression line (0.978; 95% CI 0.883 to1.073) did not differ significantly from 1 (perfect agreement): the regression formula was IMx venous = 0.751+0.978×T-MS fingerprick. Inter-patient variability calculated using the method of Becker was 42.7% (IMx venous), 50.2% (T-MS venous), and 50.4% (T-MS fingerprick).

Conclusions: Tacrolimus levels can be reliably monitored by T-MS using fingerprick samples. T-MS fingerprick levels showed a highly significant relationship to IMx venous levels and the techniques can be used interchangeably. This will allow tacrolimus levels to be collected at home and mailed to the laboratory, thus reducing parental inconvenience and improving school attendance.


H. Narchi, R. Donovan.Sandwell General Hospital, West Bromwich, UK

Aims: Children under the age of 4 years are susceptible to develop renal scarring following a urinary tract infection (UTI). Early dimercapto-succinic acid (DMSA) scintigraphy and renal power Doppler during the acute infection are useful to diagnose acute pyelonephritis changes that may predispose to permanent scarring; however, only half of these acute changes persist as permanent renal scars on repeat DMSA several months later. As scintigraphy involves exposure to radiation, strategies are needed to diminish unnecessary exposure to early as well as late DMSA scan. We investigate such a strategy of using early renal power Doppler findings to predict the children who may not be required to undergo late DMSA scintigraphy to detect renal scars.

Methods: Children <4 years of age with a first bacteriologically diagnosed UTI underwent a renal power Doppler study soon after diagnosis, in addition to standard renal tract imaging studies, including DMSA scintigraphy 6 months later to assess the presence of renal scars. The positive and negative predictive values of the early renal power Doppler (as well as its combination with ultrasound) to renal scarring on late DMSA were calculated.

Results: Twenty three children (13 females) with a median age of 30 months (including five below 12 months) were enrolled. Thirteen had a febrile presentation, including two with bacteraemia. All were kept on antibiotic prophylaxis and none had a UTI recurrence before the DMSA scintigraphy 6 months later. Permanent scarring occurred in 13% of children. In the 46 kidney units studied, initial renal power Doppler was abnormal in two and late DMSA abnormal in three units. Overall concordance between renal power Doppler and DMSA was 93.5%. The sensitivity of early renal power Doppler and of combined ultrasound and renal power Doppler for renal scar as per DMSA was respectively 33.3% and 66.7%, with an identical specificity of 97.7%; a positive predictive value for renal scarring of 50% and 66.7%; and a negative predictive value of 95.4% and 97.7%.

Conclusions: The probability of not diagnosing scarring in children <4 years if DMSA is not carried out months after a UTI is 4.6% in those with an initial normal renal power Doppler and 2.3% in those with an initial normal renal ultrasound and normal renal power Doppler.


A. T. Anbu, P. A. Smith, K. O’Donnell, G. T. N. Besley.Royal Manchester Childrens Hospital, Manchester, UK

Introduction: The starting dose of subcutaneous recombinant erythropoietin (rHuEPO) is 60 U/kg/week increasing incrementally up to 300 U/kg/week until target haemoglobin is achieved. If target haemoglobin is not achieved at this dose the patient requires review for all factors known to be associated with EPO resistance, including carnitine levels. Carnitine is a quaternary ammonium compound that helps in the stabilisation of the erythrocyte membrane and takes part in β oxidation. Studies in adults have consistently shown that carnitine deficiency is associated with higher dose requirement of rHuEPO. Carnitine supplementation has also shown to reduce the required rHuEPO dose. Previous reported studies in children do not support this.

Aims: To find out the incidence of carnitine deficiency in a cohort of children needing doses of rHuEPO >300 U/kg/week.

Methods: All children with chronic renal impairment and needing >300 U/kg/week rHuEPO (NeoRecormon® epoetin β) are included in this prospective study. Target ferritin was 100–800 mcg/l achieved with oral or intravenous iron. Target PTH was <3× upper limit of normal range (10–71 pg/ml). Renal association paediatric haemoglobin targets were used. Blood samples were taken just prior to a dialysis session in HD and at clinic visits in PD and children not yet requiring dialysis. Serum carnitine levels were analysed by Tandem Mass Spectrometry.

Results: Forty two children (42% Asian, 58% Caucasians) on PD, HD, or not yet requiring dialysis were on treatment with rHuEPO during the 3 months of the study. Ten (24%) children (ages 1–16.5 years) required a rHuEPO dose of >300 U/kg/week (range 303–631.5 U/kg/week, median of 500 U/kg/week). One child aged 3 years 5 months (on CCPD from day 10 of life) had a low free carnitine of 16.7 μM (normal range 20–40 μ). His haemoglobin was below target at 9.0 g/dl, rHuEPO dose 631.5 U/kg/week, PTH 42 pg/ml, and ferritin 546 mcg/l). Nine children had normal carnitine levels mean 40 μM (range 25.3 to 50.3 μM). Two of the nine achieved the target haemoglobin levels with a target range PTH (mean 97.6 pg/ml) and ferritin with a mean rHuEPO dose of 478.3 U/kg/week. One child achieved target haemoglobin with high PTH level and target ferritin. Six of the nine children with normal carnitines had a haemoglobin below target, mean rHuEPO dose 363.8 U/kg/week, all had PTH > target (mean 916.6 pg/ml) and target ferritin levels.

Conclusion: In this single centre study carnitine deficiency was identified in one of the 10 children requiring rHuEPO >300 U/kg/week. As previously reported hyperparathyroidism remains a much more common cause of EPO resistance, here identified in seven out of 10 children. It is definitely worth measuring carnitine levels in children requiring a rHuEPO dose of >300 U/kg/week. Prompt oral carnitine supplementation can then be started in those children identified with low levels.


J. Ellis1, A. Walker1, M. Irama1, J. Axton1, H. Bode1, R. Coward3, J. Senkungu1, D. Nansera1, D. Peat2, P. Mathieson3.1MUST, Mbarara, Uganda; 2Department of Histopathology, Norwich, UK; 3Academic renal unit, Southmead hospital, Bristol, UK

Introduction: Proteinuric renal disease is common in Africa and accounts for up to 3% of medical admissions.1 In general, a post infectious aetiology has been suspected in the majority of cases, with quartan malaria being implicated most frequently.23 HIV nephropathy is increasingly recognised in the developed world,4 but renal manifestations of HIV have not been studied extensively in Africa. We previously reported5 a retrospective analysis of admissions to the children’s ward in Mbarara University Hospital, south-western Uganda, in which we described an incidence of proteinuric renal disease that was at least eight times the UK incidence of childhood nephrotic syndrome. With full ethical approval we have performed a prospective study of proteinuric children presenting to Mbarara hospital in Uganda.

Subjects: Between March 2001 and April 2003 60 children (25 female/35 male) with greater than +2 protein on urine dip-sticking were recruited; age range 2–14 years (mean of 8.4 years). All children were clinically oedematous.

Design: Children were investigated with analysis of urinary sediment, a blood smear, C3, level, C4 level, creatinine, ASOT titre, anti DNAse B, renal ultrasound scan, and if clinically fit a percutaneous renal biopsy (analysed by a histopathologist in the UK).

Results: No patient had a blood smear suggestive of Plasmodium malariae. There were suppressed C3 and C4 complement levels in 88%. ASOT was raised in 64% and anti-streptolysin titres in 50%. Renal biopsy was performed in 39 patients and in 30 adequate tissue was obtained. 66% of these biopsies showed diffuse proliferative glomerulonephritis with prominent eosinophilic infilitration, which is rarely found in proteinuric patients in the Western world. Red cell casts were the best predictor of this finding. No biopsies suggested HIV associated nephropathy. Results from biopsies were available and communicated back to Mbarara within a week for the majority of patients, resulting in beneficial alterations in management for a number of children.

Conclusions: Our study suggests that malaria and HIV disease are not major contributors to proteinuric disease in Mbarara but another, as yet unidentified, presumed infectious trigger seems to be causing the majority of cases. The use of expensive tests such as complement levels, ASOT, and anti DNAse B levels were not helpful in predicting biopsy findings. Renal biopsies analysed in England can be processed and communicated back to rural Africa in a time span that can aid patient management.






Abstract G67