INTRODUCTION

Chronic kidney disease (CKD) in children is rare. Nevertheless, when it occurs its effects are devastating to the child’s development, generating high morbidity.

When diagnosed, most patients are in little advanced stages of the disease, in which therapeutic interventions to change its course can still be applied. In a small number of patients the disease progresses every year into established chronic kidney disease (ECKD), where the only possible treatment is replacement therapy (dialysis-transplant). Hence, early diagnosis is of paramount importance.1

The information on the characteristics of CKD in children is quite limited and has been preferably extracted from patients who were in terminal stages of the disease.2,3

The data available concern small populations in specific geographic areas, which are not always related to the general population. In the existing records, the methodology used to classify and establish a ranking of kidney diseases has not been uniform, which has hampered the comparisons of the disease characteristics in the different geographic areas.4,6

The existing registries include the European Dialysis and Transplant Association (EDTA), which currently collects data from 12 European registries of terminal patients and patients on replacement therapy (dialysis-transplant),7 the Australia-New Zealand Registry,8 the Lorraine Registry in France and the Chilean Registry, among others. The Japanese Society of Paediatric Nephrology has provided very valuable reports on CKD.

The American Registry, that is, the Registry of the North American Paediatric Renal Trials and Collaborative Studies (NAPRTCS)9 began in 1987 with the aim of collecting data from most paediatric renal transplant centres in the USA, Canada, Mexico and Costa Rica. In 1992, data of patients who were on dialysis were also included. Since 1994, all patients under 20 years of age with non-terminal CKD and GFR <75 ml/min/1,73 m2 have been included in the registry. In this context, on 31st December 2008, 7,037 patients with non-terminal CKD had already been included in the registry. Participation in this registry is voluntary, and its main disadvantage is that it only includes approximately 75% of all American patients with CKD.

In 2003, the data of the Italian registry ITALKID10 were published. This is a prospective project that includes all Italian paediatric patients with CKD. It is the most comprehensive project on the epidemiology of CKD in children. ITALKID began in 1990, and it includes a total child population of 16.8 million children.

In 2007, the Spanish Association of Paediatric Nephrology (AENP) launched the Spanish Paediatric Registry of Renal Failure (REPIR II), which is designed to include all Spanish paediatric patients with non-terminal CKD. The aims of this registry are to assess the incidence and prevalence of CKD in children, describe the natural history of the disease, explore the factors that may influence the course of the disease and develop a standard protocol for the prevention and treatment of these patients.

This paper describes the characteristics of the nascent REPIR II project and provides the first data on the incidence and prevalence of non-terminal CKD in Spanish children.

MATERIALS AND METHODS

In 2007, a committee was created within AENP to carry out the REPIR II project. This committee is composed of a chairperson, two coordinators and ten representatives of all the centres participating in the project. The number of participating centres is 46 and these are distributed across Spain, representing all the autonomous regions that make up the Spanish state. At each centre there is a paediatrician responsible for monitoring all children with kidney disease in the region concerned, who is the researcher responsible for including in the registry all patients with non-terminal CKD who meet the inclusion criteria. Furthermore, this paediatrician has the responsibility to contribute to the registry annually, by including the developmental results of each of his or her registered patients. These data should refer to the last consultation prior to 31st December of the current year.

Classification of chronic kidney disease

In this registry we have adopted the CKD classification described by the Clinical Practice Guidelines (CPG) of the National Kidney Foundation’s Kidney Disease Outcomes Quality Initiative (KDOQI guidelines)11 which classify the severity of the disease into five stages: stage 1, where the GFR is >90 ml/min/1.73 m2; stage 2 where it is between 89 and 60 ml/min/1.73 m2; stage 3 where it is between 59 and 30 ml/min/1.73 m2; stage 4 where it is between 29 and 15 ml/min/1.73 m2; and stage 5 if it is <15 ml/min/1.73 m2.

Inclusion criteria

Only children whose age does not exceed 18 years and have been diagnosed with CKD between stage 2 and stage 5 of the KDOQI guidelines classification can be entered in the registry. Following the start of the replacement therapy (dialysis or transplant), and although the data is stored in the database, the patient should be moved to another database, that is, the REPIR I, which records only those cases that are on a dialysis-transplant programme.

Since the KDOQI guidelines do not allow for including patients younger than 2 years of age, to know the stage of kidney disease of these children we extrapolated the percentage of loss of kidney function in each stage of the KDOQI guidelines, taking into account the reference values of GFR in children between 0 and 24 months of age.

To include a patient in the registry it is also necessary that the decrease in GFR is maintained for at least 3 months, thus avoiding to register patients who recover their function and do not, therefore, meet the inclusion criteria.

The assessment of GFR is performed according to the Schwartz formula12:

GFR in ml/min/1.73 m2 = K x T/ PCr

where: K is a constant whose value varies with age (in <2 years its value is 0.45; from 2 to 16 years it is 0.55; and from 16 to 21 years it is 0.70 in men and 0.55 in women), T is the length in cm and PCr is the concentration of creatinine in the blood expressed in mg/dl.

In the database it is also possible to assess the GFR using cystatin C according to the Filler formula.13 However, the criterion for inclusion of a patient is based solely on the GFR estimated with the Schwartz formula.

To define the presence of anaemia we followed the criteria of the KDOQI guidelines, which use the reference values for haemoglobin (Hb) of the NHANES III,14 considering that a patient has anaemia when his or her Hb falls below the 5th percentile for his or her age and gender.

To define the presence of arterial hypertension (HTN) we used the reference values for blood pressure (BP) of the Task Force, which were obtained from a normal population with an age not exceeding 18 years.15 We considered that a patient is hypertensive when systolic blood pressure (SBP) and/or diastolic blood pressure (DBP) are above the 90th percentile for his or her gender, age and height.

To assess the state of the calcium-phosphorus metabolism, we used the criteria recommended by the KDOQI guidelines16 for four parameters (calcium, phosphorus, Ca x P product and PTH), which should be maintained between the reference ranges listed in these guidelines, according to age and the stage of CKD.

Database

The registry has an online database designed specifically for that purpose, which can be accessed using a password. The Registry Commission within AENP manages this database. To register a patient it is necessary to have the informed consent of parents or guardians. This database complies fully with current legislation regarding privacy and data protection.

The database has four main windows. The first includes demographic data: patient’s identification code, date of diagnosis, date of data collection, date of birth, place of residence, drop-down list of diagnoses predefined on primary nephropathy (this ensures that there is uniformity in the diagnoses applied to each patient), presence of extra-renal pathology and family history.

The second window includes anthropometric and laboratory parameters: weight, height, body surface area, BMI: body mass index, head circumference (in children under 3 years of age), BUN, urea, creatinine, GFR estimated according to the Schwartz formula, GFR estimated according to the cystatin formula, creatinine clearance in 24-hour urine, BP, haematocrit (Hct), Hb, ferritin, transferrin saturation index (TSI), calcium, phosphorus, PTH and 25-(OH) vitamin D.

The third window registers the treatments received by each patient: statins, preferences of haemodialysis, patients on phosphate binders, vitamin D, 1.25-(OH) vitamin D, iron, erythropoietin, sodium bicarbonate, ion-exchange resins, sodium chloride, growth hormone, folic acid, allopurinol, nasogastric/gastrostomy, nutritional supplements, fluorhydrocortisone, hypotensive and antiproteinuric medicines.

The fourth window shows the reason and/or fate of patients who leave the database: ECKD, loss during follow-up, transfer to another centre, transfer to the adult nephrology department, death, recovery of kidney function and further loss of kidney function after having recovered it.

Finally, it should be noted that, given the short history of the REPIR II Project, this study only reflects the incidence and prevalence data of patients who were registered by the end of 2008.

RESULTS

37 centres with a total of 605 registered patients had been included by 31st December 2008. Of these patients, 74 (12%) were diagnosed with CKD in 2008. The average age of all patients at diagnosis was 3.9 ± 5 years, while at the time of the last consultation it was 9.2 ± 6 years. 59%, 13%, 22% and 6% of the patients had an age at diagnosis of <2, 3-5, 6­12 and 13-18 years, respectively. However, during the last consultation for the same age groups, the percentage was 18%, 18%, 34% and 31%, respectively. The difference between the average age at the last consultation and at diagnosis is high (± 5 years), indicating that a significant number of patients registered have already a prolonged evolution of several years. 66% of the patients (400 cases) were male and the rest were female (205 cases, 34%). The quotient of men and women was 1.95. 88% of the patients were Caucasian, 6% Hispanic, 3% Arabs, 1% Black and 2% were of other races.

At the time of the study, in 2008, on a total Spanish population of 46,157,822 million inhabitants and a population of 11,296,170 million with an age equal to or less than 18 years, the incidence and prevalence of non-terminal CKD was

8.66 ppm and 71.06 ppm under the age of 18 years, respectively.

Primary nephropathy

Structural abnormalities were the most common primary nephropathy (356 patients, 59%). There were 19 cases (3%) with a glomerulopathy, 86 (14%) with a hereditary or cystic disease, 67 (11%) with vascular disease and 75 (12%) with other nephropathies. Table 1 shows the primary nephropathies in relation to the age of the patients. It is evident that in all age groups the prevalence of structural abnormalities remained the same while the number of glomerulopathies was small. 76% of the structural abnormalities occurred in men while glomerulopathies were more common in women (58%). The remaining kidney diseases were distributed equally in both genders.

Renal function

Table 2 shows the renal function in all patients together at the time of the last consultation in 2008 and distributes them according to each age group and stage of the disease. The mean GFR of all patients was 52 ± 2 ml/min/1.73 m2. 82% of all patients were in stage 2 or 3, that is, the CKD of patients registered up to that point in time was mild-moderate.

Anaemia

The mean values for Hct, Hb, ferritin and TSI when assessed for all patients together was 37.5%, 13 ± 2 g/dl, 91 ± 109 ng/ml and 25 ± 14%, respectively. In addition, 30% of patients had anaemia. As the disease progressed, the percentage of anaemia increased, reaching 14%, 33%, 58% and 54% in stages 2, 3, 4 and 5, respectively. The presence of anaemia in different age groups was irregular and occurred in 31%, 19%, 26% and 39% of children aged <2, 3-5, 6-12 and 13-18 years old, respectively.

Table 3 shows the anaemic and non-anaemic patients according to age and treatment received. 33% (57 patients) of the anaemic patients did not receive any treatment and 67% (117 patients) were inadequately treated with iron and/or erythropoietin. 34% of non-anaemic patients (139 patients) received some type of treatment.

Arterial hypertension

Table 4 shows the prevalence of HTN in registered patients with regard to age groups and the stage of the disease. Notably, the overall percentage of patients with HTN was 19%, and its presence was not related to age or stage of the disease. 63% of glomerulopathies, 31% of hereditary and cystic diseases, 20% of vascular diseases and 16% of the group of patients with other diseases had HTN. The aetiology group, which had the smallest number of hypertensive patients, was the one with structural abnormalities (14%).

6% of patients with HTN (7/114) were not receiving any hypotensive treatment. ACE inhibitors and/or ARBs were the most used medicines in patients who received treatment (71% of the patients treated). 59% received a single medicine, 33% 2 medicines and 8% 3 or 4 medicines, with an average of 1.52 medicines/patient.

Phosphocalcic metabolism

The average values of calcaemia, phosphataemia, Ca x P product and PTH in all patients evaluated together were 9.8 ±

0.6 mg/dl, 4.7 ±1.0 mg/dl, 46 ± 10 mg/dl and 104 ± 113 ng/ml, respectively.

Table 5 shows that when the assessment is made based on age, calcaemia does not change significantly. Phosphataemia and the Ca x P product present a slight decrease with age. The PTH remains the same without significant differences in all age groups, except for the group between 0-2 years of age in which it is higher.

The same does not happen when the assessment of these parameters is carried out according to the stage of the disease. Table 5 shows that while calcaemia remains normal and similar in almost all stages, phosphataemia, the Ca x P product and the PTH rise significantly as the severity of the disease increases. 17% of the patients complied with the KDOQI recommendations on phosphocalcic metabolism, keeping these four parameters within the recommended ranges. 35% complied with three of the recommendations, 34% with two, 8% with one and 1% did not comply with any recommendation.

Growth

The overall mean of the size Z-Score in all patients in the last consultation was -1.03 ± 2. The growth deficit was inversely related to the patients’ age. Table 6 shows that the age group with the worst growth was 0-2 years. On the other hand, the more the disease advanced, the worse the growth became, with the patients in stage 5 having the worst growth rate.

However, when performing a multivariate logistic regression analysis, in which various factors were analysed (age, gender, race, primary nephropathy, CKD stage, presence of anaemia, presence of acidaemia, value of PTH and value of the Ca x P product), only age significantly predicted patients with short stature. All the children under 2 years old had a 40% higher probability of having short stature (OR = 1.40; P <0.01).

136 patients (25%) had a size Z-Score of -1.88 or worse. It should be noted that within this group, 82% of the patients were not treated with growth hormone.

7% (42 patients) of the patients were malnourished, with a BMI Z-Score of -1.88 or worse. In addition, most of them were in the 0-2 years age group. That is, the children under two years of age were those with poorer growth and the most malnourished.

DISCUSSION

This is the first registry that makes a prospective analysis of the incidence, prevalence, aetiology and comorbidity of CKD in the Spanish paediatric population.

The natural history of kidney disease is variable and at times unpredictable. However, in children, there is usually a continuing deterioration, which often leads to ECKD. The progression of kidney disease is greater during the two periods of rapid growth, that is, early childhood and puberty, in which the sharp increase in body mass results in an increase in the demand for filtration of the remaining nephrons. Hormonal changes during puberty contribute to the rapid deterioration of glomerular filtration, and may also influence genetic, family, or ethnic factors. Finally, other factors associated with the progression of CKD are also present in adulthood: AHT, proteinuria, obesity, dyslipidaemia, anaemia or alterations in the phosphocalcic metabolism.17

In 2008, the incidence of non-terminal CKD in our centres was 8.66 ppm of the population under 18 years of age, while the prevalence was 71.06. These figures are very close to those mentioned in the Italian registry in 2003 (12.1 and

74.07 ppm <18 years of age of incidence and prevalence, respectively).10 However, given the short history of the REPIR II project, these data are provisional, and in the coming years, they will likely undergo changes that could become important.

As in all paediatric series, the structural pathology, and therefore the congenital pathology, is the leading cause of CKD, reaching 59% in our series, 56% in the American registry and 67.5% in the Italian one. This results in the age at diagnosis to be under 6 years old in most cases (72% in our registry). The fact that the difference between the average age at the last consultation in 2008 and the age at diagnosis of CKD is high (± 5 years) indicates that a significant number of patients registered in our database now have a prolonged evolution of several years. On the other hand, given the characteristics of the Spanish population, it is comprehensible that we found a clear predominance of Caucasians in our series (88%).

The percentage of glomerulopathies as a cause of non-terminal CKD shown in Table 1 is very small (3%) and contrasts with the one that occurs in ECKD, which may reach 20%. This is because the evolution of glomerulopathies is much faster than that of renal dysplasias, many of which reach ECKD in adulthood.18

The mean value of GFR in our patients is 52 ± 2 ml/min/1.73 m2, with 82% in stages 2 and 3, whereas in the NAPRTCS and the ITALKID it is 39 and 42 ml/min/1.73 m2, respectively. That is, the patients in our registry have a milder CKD than those of the other two registries (Table 2). This may be related, at least in part, to the fact that paediatric patients with CKD are referred in our centres much earlier (from stage 1) from primary care services to paediatric nephrology departments.

The information available in the medical literature on the comorbidity associated with CKD is very small, while only the NAPRTCS offers annual data on this subject. In this study, we examined the prevalence of anaemia and HTN, as well as the level changes in the calcium-phosphorus metabolism and the growth that occurs in these patients.9,16,19

Anaemia is a common complication in advanced CKD and is associated with an increase in morbidity and mortality 18,20,21 The prevalence of anaemia in our patients was 30%. This percentage increased with the severity of CKD, in such manner that in the pre-dialysis stages 4 and 5, it was 58% and 54%, respectively. We found that our anaemic patients were being treated improperly, and that 33% of them received no treatment while the rest received insufficient treatment (Table 3).

Arterial hypertension is a known factor of cardiovascular risk and CKD progression in adults, and although there are no prospective data on children,22,23 it is logical to think that it is also true in children. In our series, only 19% of our patients had arterial hypertension. This percentage was much lower than that of the American registry, which concerns a 48% prevalence of hypertension.9 Although this datum has to be confirmed in the coming years, two of the factors that could explain it would be that the most common aetiology among our patients was structural disease, which is not usually associated with HTN, and that most children had mild to moderate kidney disease. On the other hand, at present, we cannot exclude that we have under-diagnosed cases of HTN, since there are indications that there was no unanimity among the various participating centres when defining HTN (Table 4). We only found a relationship between the prevalence of HTN and the aetiology of CKD, in such manner that 63% of the children with glomerulopathy had HTN. We were unable to show an association with age or the stage of CKD. 94% of patients with hypertension received hypotensive treatment and 71% of them were treated with ACE inhibitors and/or ARBs.

Abnormalities in the bone mineral metabolism are a universal finding as they progress to CKD. In children, they also have a deleterious effect on growth.24,25 Although in our patients the mean values for calcaemia, phosphataemia, the Ca x P product and PTH, which were valued together, were within the limits of the recommendations of the KDOQI guidelines,26-28 when we analysed the same parameters taking into account the age and the stage of the disease, only 17% met the four recommendations, while most only met three (35%) or two (34%) recommendations. Similar to other authors, we found that the recommendations of the KDOQI guidelines on the management of changes in the Ca-P metabolism are difficult to fulfil (Table 5).

Growth retardation is a common problem in children with CKD and it is associated with a significant morbidity and mortality rate.18 The final height of children with early-stage CKD is one of the most important factors in their rehabilitation.18,29-31 The overall mean of the size Z-Score at the last consultation of our patients was -1.03 ± 2. There were 136 patients (25%) who had a size Z-Score of <1.88, a figure that contrasts with those of the American registry, where the number of patients with growth retardation was higher (36%). On the other hand, we found that the growth deficit was inversely related to age and the severity of CKD, presenting a worse growth in younger patients (0-2 years age group) and in those who were in stage 5 pre-dialysis (Table 6). However, when performing a multivariate logistic regression analysis, only age significantly predicted patients with short stature. All the children under 2 years old had a 40% higher probability of having short stature (OR = 1.40; P <0.01).

The percentage of malnutrition (BMI Z-Score <1.88) in our series was 7%, mostly in the 0-2 years age group. That is, the children under two years of age were those with poor growth and the most malnourished.

To conclude, this paper presents the first data from 2008 of the Spanish Registry on non-terminal CKD in children from the REPIR II project. This is the first study that performed a prospective analysis of the incidence, prevalence, aetiology and comorbidity of CKD in the Spanish paediatric population. Given the short history of this registry, the data presented are provisional and may undergo significant changes in the coming years.

Acknowledgements

This database is funded by Pfizer. We thank Francisco Camacho Poyato, Medical Advisor at Pfizer, for his unconditional support for this project, without which this study could not have been carried out. We want to extend our gratitude to all the paediatricians and paediatric nephrologists who, with their input of patient data, have enabled us to begin to understand the characteristics of CKD in Spanish children.

10402_108_8365_en_10402_tabla_1_en.doc

Table 1. Primary nephropathy at diagnosis

10402_108_8366_en_10402_tabla_2_en.doc

Table 2. CKD stage at the time of the last consultation in relation to patient age

10402_108_8367_en_10402_tabla_3_en.doc

Table 3. Presence of anaemia at the time of the last consultation in relation to age and treatment received

10402_108_8368_en_10402_tabla_4_en.doc

Table 4. Frequency of hypertension at the time of the last consultation in relation to age and severity of CKD

10402_108_8369_en_10402_tabla_5_en.doc

Table 5. Phosphocalcic metabolism: mean and standard deviation values of the various parameters at the time of the last consultation in relation to age and severity of CKD

10402_108_8370_en_10402_tabla_6_en.doc

Table 6. Patient growth the at the time of the last consultation in relation to age and severity