This document summarizes recent advances in identifying urinary biomarkers for pediatric renal diseases. It discusses how proteomic techniques have led to the identification of urine as a good source for non-invasive biomarkers. Several studies identified potential biomarkers for ureteropelvic junction obstruction and renal Fanconi syndrome in children, showing promise for urinary proteomics in pediatrics. However, more research is still needed using appropriate validation methods to identify clinically useful biomarkers for other pediatric renal diseases.

1. REVIEW
Advances in urinary proteome analysis and biomarker
discovery in pediatric renal disease
Cécile Caubet & Chrystelle Lacroix &
Stéphane Decramer & Jens Drube &
Jochen H. H. Ehrich & Harald Mischak &
Jean-Loup Bascands & Joost P. Schanstra
Received: 29 April 2009 /Revised: 1 June 2009 /Accepted: 2 June 2009 /Published online: 15 July 2009
# IPNA 2009
Abstract Recent progress in proteomic analysis and strate-
gies for the identification of clinically useful biomarkers in
biofluids has led to the identification of urine as an excellent
non-invasive reservoir for biomarkers of disease. Urinary
biomarkers have been identified and validated on indepen-
dent cohorts in different high-incidence adult renal diseases,
including diabetic nephropathy, chronic kidney disease and
immunoglobulin A-nephropathy, but also in extrarenal
disease, such as coronary artery disease. Unfortunately, this
type of research is underrepresented in the pediatric
population. Here, we present the rare studies in the pediatric
population that identified potential clinically useful urinary
biomarkers in ureteropelvic junction (UPJ) obstruction and
renal Fanconi syndrome. These studies, although limited in
number, clearly show the potential of urinary proteomics,
especially in the pediatric population. It is anticipated that the
advances made in the adult population, the lessons learned
on the use of appropriate statistics and the inclusion of
independent blinded validation cohorts in these types of
studies will rapidly lead to clinical useful urinary biomarkers
for other pediatric (renal) disease in a population where non-
invasive analysis is particularly appreciated.
Keywords Biomarkers . Fanconi syndrome . Proteomics .
Statistics . Ureteropelvic junction obstruction . Urine .
Validation
Biomarkers in biofluids: from blood to urine
For several decades biofluid biomarkers have been playing
an important role in diagnosing various diseases and
disease stages. However, until recently, the identification
of novel markers has been an arduous task. This has
changed dramatically with the development of high-
throughput proteomic techniques for screening biofluids
which has enabled many potential different biomarkers to
be assayed simultaneously. The results of such screening
assays has revealed that many diseases cannot be described
by a single biomarker but, rather, by a panel of several
biomarkers. In 2002, Petricoin and colleagues were the first
to identify proteomic patterns in serum for the identification
C. Caubet :S. Decramer :J.-L. Bascands :J. P. Schanstra (*)
Institut National de la Santé et de la Recherche Médicale
(INSERM) U 858-I2MR-Equipe no. 5,
1 avenue Jean Poulhès, BP 84225,
31432 Toulouse, Cedex 4, France
e-mail: joost-peter.schanstra@inserm.fr
C. Caubet :S. Decramer :J.-L. Bascands :J. P. Schanstra
Institut de Médecine Moléculaire de Rangueil,
Equipe no. 5, IFR150, Université Toulouse III Paul-Sabatier,
Toulouse, France
C. Lacroix
Institut de Pharmacologie et de Biologie Structurale (IPBS),
CNRS, Toulouse, France
C. Lacroix
UPS, IPBS, Université de Toulouse, Toulouse, France
S. Decramer
Department of Paediatric Nephrology,
Centre de Référence du Sud Ouest des Maladies Rénales Rares,
Hôpital des Enfants, Toulouse, France
J. Drube :J. H. H. Ehrich
Department of Paediatric Kidney,
Liver and Metabolic Diseases, Children’s Hospital,
Hannover Medical School, Hannover, Germany
H. Mischak
Mosaiques Diagnostics and Therapeutics AG,
Hannover, Germany
Pediatr Nephrol (2010) 25:27–35
DOI 10.1007/s00467-009-1251-5

2. of ovarian cancer [1]. This study attracted massive interest
from both the clinical and research community. However,
the initial optimism generated by this research was rapidly
dampened by follow-up studies showing that the results of
this study were irreproducible [2], most likely due to the
improper mass calibration of the mass spectrometer,
technical flaws in the experimental design and improper
execution of the experimental protocol. Concomitantly,
there has been active discussion on whether blood is a
good source of biomarkers for disease, as blood collection
is inevitably associated with the activation of proteases.
These generate an array of proteolytic breakdown products
and introduce substantial variability, although some studies
used protease activity to define disease states [3-5]. Further,
a very few proteins constitute 99% of the total blood
proteins, thus blocking the efficient identification of the less
abundant proteins. The removal of these few but abundant
proteins is not 100% efficient and also introduces additional
variability during sample preparation [6].
While the interest for blood as a source of biomarkers
was fading, urine emerged as a potential and more suitable
reservoir for identifying biomarkers. In contrast to blood,
the pre-analytical handling is simple, and urine has been
proven to be particularly stable [7, 8]. Both of these factors
significantly reduce the variability of the samples and thus
favor the discovery of disease biomarkers. Urine has the
disadvantage that it shows a wide variation in protein and
peptide concentrations, mostly due to differences in the
daily intake of fluid. However, this shortcoming can be
countered by standardization based on creatinine [9] or
peptides generally present in urine [10]. The urinary
proteins and peptides are of different origin and include
filtered and secreted plasma proteins, proteins secreted by
various renal segments, proteolytic degradation products of
extracellular matrix, proteins secreted by the urinary tract
and proteins derived from dead shedded cells along the
nephron and the urinary tract. Under physiological con-
ditions, around 70% of the urinary proteins are estimated to
be derived from the kidney and the urinary tract [11]. For
these reasons, urine is an interesting source of biomarkers
to determine the health status of both the kidney and
extrarenal organs where biomarkers transported by blood
are filtered or secreted into the urine [12].
Tools and strategies to study the urinary proteome
and identify biomarkers
The study of the urinary proteome has become possible by
the significant technological advances in mass spectrometry
and profiling techniques over the last few years. Almost all
known mass spectrometry techniques have been used for
the analysis of the urinary proteome, including two-
dimensional gel-electrophoresis followed by mass spec-
trometry (2DE–MS), liquid chromatography coupled to
mass spectrometry (LC–MS), surface-enhanced laser de-
sorption/ionization coupled to mass spectrometry (SELDI–
TOF) and capillary electrophoresis coupled to mass
spectrometry (CE–MS). Detailed comparison of these
different techniques can be found in recent reviews [13,
14]. All of these techniques employ pre-fractionation to
reduce the complexity of the samples. This step can consist
of the selective absorption of proteins and peptides with
similar physicochemical characteristics on a surface
(SELDI), electrophoretic separation (capillary, 2D-gel) or
liquid chromatography (Fig. 1). The obtained fractions are
ionized and introduced into a mass spectrometer where the
mass and abundance of the proteins and peptides are
recorded. All of these different techniques enable analysis
of the urinary proteome, and each has its own distinct
advantages and disadvantages (Table 1).
Special attention should be paid to basic analytical
principles in order to guarantee a high grade of validity
and reproducibility of clinical application of the identified
biomarkers. This issue has been discussed in detail in a
number of recent papers [15-17]. The following factors play
a crucial role: (1) a single and clear clinical question, (2) a
large number of urine samples obtained in a standardized
fashion in the test and control group, (3) analysis by
instrumentation allowing relatively high throughput and
high reproducibility, (4) appropriate statistical analysis for
large sample numbers (correction for multiple testing) and
(5) validation of the potential biomarkers in a blinded study.
The fourth and fifth factors mentioned above are of
critical importance. The reasons for this are detailed below:
(4): The assessment of statistical validity in the absence of
multiple testing is inappropriate and misleading, but
unfortunately still widely used. This subject, which
represents an issue for all multiparametric approaches,
such as genomics, metabolomics or proteomics, has
been discussed for the proteomics field in detail in a
recent review [18]. In a recent experiment involving the
definition of gender-associated biomarkers, we were
able to demonstrate that even the distribution of “true
significant biomarkers” (biomarkers that were found to
be significantly associated with gender in the indepen-
dent blinded test set) is similar between the groups of
“apparently significant biomarkers” (having an unad-
justed p value <0.05) and “apparently insignificant
biomarkers” (having an unadjusted p value >0.05). The
fraction of “true significant biomarkers” was essential-
ly identical in both groups, further demonstrating that
the unadjusted p value does generally not provide any
information in a typical multiparametric experiment
(Harris et al., in preparation).
28 Pediatr Nephrol (2010) 25:27–35

3. (5): Underlying and generally unknown bias as well as
unavoidable biological variability in the samples
analyzed generally result in the identification of
potential biomarkers (based upon correct statistical
assessment) that are in fact not associated with the
investigated (patho)physiological condition. Conse-
quently, validation of the potential biomarkers in an
independent test set is mandatory. What is more:
machine learning tools, such as support vector
machines, artificial neural networks or others that
are used to combine several biomarkers into a multi-
marker model, frequently tend to “overfit” data [18,
19]. This overfitting results in excellent classification
of the training set (even 100% accuracy can be
achieved) but, at the same time, the model only
applies to the training set and completely fails to
correctly classify additional datasets. As a conse-
quence, the testing of both defined biomarkers and, if
applicable, the established biomarker model on an
independent masked/blinded set of samples large
enough to show statistical significance appears to be
mandatory. The p value should be <0.05, and if
biomarker models are established, the area under the
curve (AUC) in the receiver operating characteristic
(ROC) analysis should at least be >0.7. In the absence
of such data, the validity of the reported results cannot
be assessed, rendering them essentially meaningless.
The potential use of a protein or peptide as a biomarker
depends on how selective and sensitive it enables the
Table 1 Advantages and disadvantages of proteomic platforms that can be used in urinary biomarker discovery
Technology Advantages Disadvantages
2DE–MS Large molecules can be detected and enables estimation
of actual molecular weight, sequencing of biomarkers
easy to perform from 2D spots
Small molecules (<10 kDa) not detected, difficult to
automate, time consuming, medium throughput,
moderate comparability
SELDI–TOF High throughput, easy-to-use, automation, low
sample volume
Restricted to selected proteins, low resolution MS,
lack of comparability, sensitive toward interfering
compounds.
LC–MS Automation, multidimensional, high sensitivity, used for
detection of large molecules (>20 kDa) after tryptic
digest, sequence determination of biomarkers provided
by MS/MS
Reassembly of tryptic peptides into their precursor
molecule can be problematic, time consuming,
relatively sensitive toward interfering compounds,
medium throughput
CE–MS Automation, high sensitivity, fast, low sample volume,
multidimensional
Generally not suited for larger molecules (>20 kDa)
2DE–MS, Two-dimensional gel-electrophoresis followed by mass spectrometry; LC–MS, liquid chromatography coupled to mass spectrometry,
SELDI–TOF, surface-enhanced laser desorption/ionization coupled to mass spectrometry; CE–MS, capillary electrophoresis coupled to mass
spectrometry
Fractionation Mass spectrometrySample
2D-PAGE SELDI
Capillary
electrophoresisLiquid
chromatography
Proteomes
1 2 3
Fig. 1 Proteome analysis of urine requires fractionation to reduce
complexity of the sample. 1 Fractionation can be obtained by different
chromatographic techniques or by the specific absorption of a set of
proteins on a surface. 2 These fractions are subsequently analyzed by
a mass spectrometer (MS) where the relative abundance of the
different proteins and peptides is determined. 3 Informatics treatment
of the protein data in combination with the fractionation (example:
migration time on a capillary or liquid chromatography column)
parameters yields protein profiles representing the (partial) protein
content of samples. SELDI Surface-enhanced laser desorption/ioniza-
tion, 2D two dimensional, PAGE polyacrylamide gel electrophoresis
Pediatr Nephrol (2010) 25:27–35 29

4. assessment of the disease. Most of the traditionally used
biomarkers have been identified on the basis of empirical
knowledge of the underlying disease. In general, these
single biomarkers only display moderate diagnostic
value, mostly due to low specificity. For example,
prostate specific antigen (PSA) is widely used as a
marker for prostate cancer. Its prognostic relevance,
however, is the subject of ongoing debates due to a lack
of specificity when PSA levels are only moderately
increased (4–10 ng/mL) [20]. Another example is the use
of microalbuminuria as an early non-invasive marker of
renal damage. Microalbuminuria can be present in diabetic
patients before apparent damage to glomerular function or
increased serum creatinine levels [21, 22]. However,
microalbuminuria is also found intermittently in apparently
healthy individuals and cannot be utilized with sufficient
confidence as a predictive marker of renal disease [23].
These two examples underline the need for more accurate
biomarkers. This raises the question of whether a single
marker can actually fulfill the requirements to (1) reliably
detect a disease as early as possible, (2) unambiguously
distinguish a specific disease from other pathological
conditions and (3) monitor the efficacy of therapy. An
alternative strategy is the identification of several markers
which as stand-alone markers do not present high specificity
and sensitivity but which, as a panel (or pattern), work in
concert to give high accuracy [24]. A similar approach is
used by clinicians in diagnosing a disease entity– several
symptoms and signs will eventually lead to the final
diagnosis. The general criteria that are applied to biomarkers
to be used for clinical assessment (e.g. known identity,
reproducible detection, known deviation) also apply for the
single biomarkers that make up the multi-marker panel [16].
Although not essential for the establishment of valid
signature patterns if reliable methods for definition and
detection are available (e.g. accurate mass and migration
time), it is important that the biomarkers be identified.
This is necessary from the aspect of increasing our
biological knowledge about disease processes and also in
terms of subsequent measurement using other technolo-
gies [8, 25]. Currently, the majority of commercial
diagnostic assays are immuno-capture based, and it is
very likely that any translation of the biomarkers will
involve a similar format, whether the readout involves
classical enzyme-linked immunosorbent assay (ELISA),
multiplexed immunoassays or immuno-MS. Here, we
want to emphasize that the analysis of single biomarkers
with immunological technologies requires probes that are
specific not merely for the native protein from which the
biomarker is derived, but also for the distinct biomarker
that has a defined C and N terminus as well as (frequently)
post-translational modifications. Ignoring these features
may lead to false-positive results, which must be avoided.
Use of urinary proteome analysis for biomarker
discovery in pediatric renal disease
The main focus for urinary biomarkers of renal disease is
the adult population [13, 14], in part due to the rising
prevalence of chronic kidney disease in the aging popula-
tion. However, the main scope of this review is the progress
that has been made in terms of identifying urinary
biomarkers of pediatric renal disease. For the reasons
outlined above, only studies with independent identification
and validation cohorts will be discussed herein. In addition,
although urinary proteome analysis will—over the long
term—also provide information on the etiology and patho
(physiology) of the underlying disease, we will not discuss
this issue as it is beyond the scope of our review. The reader
is referred to [13] for more information on this topic.
Ureteropelvic junction obstruction
Antenatal screening detects fetal hydronephrosis in around one
out of 100 births, with about 20% of the cases being clinically
significant. Ureteropelvic junction (UPJ) obstruction is found in
40–50% of these clinically significant cases [26]. Although
UPJ obstruction in the majority of the cases is not considered
to be a severe disease, it requires invasive follow-up.
Ureteropelvic junction obstruction is functionally defined as
a restriction to the urinary outflow that, when left untreated,
will cause progressive renal deterioration. Alternatively, this
obstruction has been more generally defined as a condition
that hampers optimal renal development [27]. Since hydro-
nephrosis is not always synonymous with obstruction, the
differentiation between a dilated obstructed and dilated non-
obstructed kidney is often a challenge, and non-invasive
techniques for assessment are needed. No such generally
accepted reference standards are currently available to
correctly identify obstruction, and the diagnosis is mostly still
based on arbitrary threshold values and the results of various
radiologic investigations that are often repeatedly performed.
Some of these imaging techniques expose these infants to
radiation and may need the injection of radiocontrast or
radioisotope material. The period of surveillance of UPJ
obstruction patients can take up to 4 years. A retrospective
study on 343 children with UJP obstruction showed that half
of the patients needed surgery; of these, 50% were operated
before the age of 2 years while the remaining 50% were
operated on between 2 and 4 years of age [28]. Consequently,
attempts have been made to use urinary proteome analysis and
identify biomarkers in infants with UPJ obstruction to predict
the need for surgical intervention at an early stage [29, 30].
In these studies, two different cohorts of UPJ obstruction
patients were employed: one for the identification and one for
the validation of urinary biomarkers of UPJ obstruction. For
the identification of biomarkers, urine samples were obtained
30 Pediatr Nephrol (2010) 25:27–35

5. before 1 month of age from healthy controls (n=13), UPJ
obstruction patients with low level obstruction (grade 1/2
hydronephrosis, as defined by [31] modified by [32], pelvic
dilatation 5–15 mm, n=19) and UPJ obstruction patients
scheduled for pyeloplasty (grade 3/4 hydronephrosis, pelvic
dilatation >15 mm, differential renal function <10% and a
washout pattern in diuretic renography with eliminated
activity at 30 min >30%, n=19). Using CE–MS for
analyzing the urinary proteome, 53 urinary biomarkers were
identified that classified these three different groups with
high specificity and sensitivity. The 53 biomarkers were then
used to predict the fate of an independent test set of 36 UPJ
obstruction patients with intermediate UPJ obstruction
(clinical characteristics between mild and severe UPJ
obstruction). In this blinded prospective study, the clinical
outcome was predicted with 95% accuracy 9 months in
advance [30]. After 15 months of follow-up, the accuracy of
the prediction increased to 97% as one of the newborns with
UPJ obstruction had to be operated at a late stage, as
predicted by the urinary proteome analysis [29]. The results
of this French study are supported by an unpublished
separate study which was performed in a German center
using slightly different criteria for need of surgery. This
study also revealed that the accuracy of the urinary proteome
pattern was restricted to the infant age. These encouraging
data resulted in the initiation of a multi-center prospective
study on 358 UPJ patients for validation of the predictive
value in independent pediatric units. The results of this
international study are expected in 2011.
Once multi-center validation has been obtained, urinary
proteome analysis may replace (at least partially) the invasive
follow-up of UPJ obstruction patients. In addition to this gain
in patient comfort, a recent assessment showed that urinary
proteome analysis can also significantly contribute to the
reduction of costs for the follow-up of UPJ obstruction [33].
The Markov process decision tree model compared the
current strategy (watchful waiting with serial imaging
overtime) with a strategy incorporating a urine proteome
analysis at birth as a marker of disease progression. The
analysis included the cost of surgery, imaging and office
visits based on hospital charge data. A total of 53 variables
were analyzed. The conclusion of this study was that the
incorporation of urinary proteome analysis in the initial
evaluation of UPJ obstruction significantly reduced costs and
increase the quality adjusted life years (QALY) in this patient
population. Incorporating the urinary proteome analysis
increased the cost-effectiveness by $8,000 per QALY per
patient [33].
Renal Fanconi syndrome
The renal Fanconi syndrome (FS) is characterized by renal
glucosuria, loss of electrolytes, bicarbonate and lactate,
generalized hyperaminoaciduria and low-molecular-weight
proteinuria. Renal Fanconi syndrome is a constellation of
laboratory findings displayed by many different inherited
diseases [34] or due to a multitude of exogenous agents,
such as antibiotics, antiviral agents, chemotherapeutics,
bisphosphonate, aristolochic acid (contained in some
Chinese herbs [35]), valproate [36] and immunosuppres-
sive, antiviral and X-ray contrast agents [37, 38]. The
diagnosis of FS is based on the analysis of urine to detect
glucosuria and low-molecular-weight proteinuria, serum
analysis and clinical examination. The proteins well known
to be excreted in FS are neither the cause nor are they
specific to distinct tubular damage as these proteins are
freely filtered in the glomerulus and not reabsorbed by
defect tubular cells.
In a small-scale study which involved the use of CE–MS
to study seven pediatric patients with cystinosis and six
patients with ifosfamide-induced FS as the patient study
group and 54 healthy volunteers and 45 patients suffering
from other renal diseases as controls, Drube et al. [39] were
able to establish a urinary proteome pattern. This FS pattern
was validated by a blinded analysis consisting of 11 FS
patients and nine patients with renal disease other than FS.
Reduced amounts of fragments of the marker proteins
osteopontin and uromodulin were found in the urine of FS
patients, indicating the loss of function of tubular excretion
in all patients regardless of the underlying cause of FS. In
addition, six different fragments of the collagen alpha-1 (I)
chain were either elevated or reduced in the urine,
indicating a change in the composition of the proteases
involved in collagen degradation, as is also observed in
interstitial fibrosis. These changes were prominent irre-
spectively of the stages of FS. This finding indicates that
fibrosis is an early starting pathogenic process for the
development of renal insufficiency in FS patients.
The specificity of urinary proteomics for detecting FS
was 89% and sensitivity was 82% The proteome pattern
established in this study using CE–MS suggests a number
of future applications in clinical medicine, such as the
routine diagnosis of renal comorbidity in children with
cytotoxic treatment of malignancies. In fact, acquired FS
was reported to occur in up to 56.7% of patients during
cytotoxic therapy in cancer treatments involving the use of
ifosfamide [40] or other cytotoxic agents. Of those pediatric
patients treated with ifosfamide, 88% developed transient
glucosuria [41], while the percentage of those retaining
renal impairment ranged from 1.3 to 27% of treated patients
[42, 43]. The development of symptoms is slow and,
consequently, FS was usually diagnosed only several
months after cytotoxic therapy [44]. A sufficiently reliable
and routine test is therefore needed to detect patients with
FS before they suffer from renal insufficiency or secondary
illnesses, such as renal rickets [44]. This study supports the
Pediatr Nephrol (2010) 25:27–35 31

6. finding of Cutillas et al. [45]. However, it remains to be
studied to what extent urinary proteome analysis may (1)
differentiate different types of hereditary and acquired
tubulopathies [46] and (2) predict progression of renal
dysfunction in FS.
Age affects the urinary proteome
The identification of urinary biomarkers of (renal) disease
in the adult population is much more advanced than that in
the pediatric population. For example, in high-incidence
diseases, such as diabetic nephropathy, urinary biomarkers
have been identified and validated on independent adult
cohorts ([24, 47–50], and see below). Therefore, if one
could exploit biomarkers of diabetic nephropathy identi-
fied in the adult population in the pediatric population
there would be a significant gain of time in the discovery
phase. The main obstacle for using adult biomarkers in the
pediatric population is the age dependence of urinary
proteome patterns in healthy infants, toddlers, children
and adolescents. In one study, the low-molecular-weight
urinary proteome of 324 healthy individuals ranging from
2 to 73 years of age was analyzed by CE–MS [51]. Age-
related modification of the secretion of 325 of the more
than 5000 urinary peptides studied was observed. Inter-
estingly, the majority of these changes were associated
with renal development before and during puberty, while
49 peptides were related to aging in adults. A substantial
fraction of these aging-related peptides were also markers
of chronic kidney disease and scored particularly well
with diabetic nephropathy. In fact, 22% of the urinary
peptides associated with aging had also previously been
identified as urinary biomarkers of diabetic nephropathy.
Two additional observations were made in this study: (1)
the identification of aging-related peptides suggested the
involvement of reduced proteolytic activity in older
patients, thus correlating human data with that of animal
experiments, and (ii) a number of the 324 supposedly
healthy individuals had a urinary peptide pattern suggest-
ing an individual significantly older than his/her actual
age. Similar studies on the aging renal transcriptome also
identified some outliers and confirmed, on a histological
level, the presence of renal lesions in supposedly healthy
individuals [52]. While more work needs to be done,
urinary proteome analysis may allow clinicians to non-
invasively pinpoint individuals in the aging population
that appear to suffer from yet clinically unapparent
cardiovascular and kidney damage.
In the near future, this database of the modification of
the urinary proteome with aging in combination with the
existing database of low-molecular-weight urinary markers
of a variety of renal diseases [53] will allow testing of the
hypothesis that adult biomarkers, corrected for age based on
the known proteomics differences, can be used in the
pediatric population (and vice versa).
Urinary biomarker discovery in high-incidence adult
renal disease
The incidence of type II diabetic nephropathy (DN), long
reserved for the older population, is currently also rising in
the pediatric population [54], and similar tendencies have
been observed for type I diabetes [55-57]. In the adult
population, DN has become the most prevalent cause of
end-stage kidney disease and is the most common and
serious complication of both type I and type II diabetes,
affecting up to 40% of all diabetic patients [58]. Currently,
the best predictor of progression to DN is the low-grade
elevation of urinary albumin excretion (UAE) between 30
and 300 mg/day (microalbuminuria) at which time various
degrees of renal structural damage may be present but
where renal function is usually within normal levels.
Additional risk factors for progression to DN are increased
arterial pressure and poor glycemic control, but these only
explain a minor fraction of the total risk for developing DN.
More specific and sensitive risk markers are needed to
identify the high-risk individuals in the diabetic population.
In one study on 305 individuals, biomarkers for DN
were defined and validated in blinded data sets using
CE–MS [47]. A panel of 40 biomarkers distinguished
patients with diabetes from healthy individuals with 89%
sensitivity and 91% specificity. Among the patients with
diabetes, 102 urinary biomarkers differed significantly
between patients with normoalbuminuria and nephropathy,
and these allowed the authors of the study to construct a
model that correctly identified diabetic nephropathy with
97% sensitivity and specificity. This study presented two
additional interesting features in that these biomarkers (1)
also identified patients with microalbuminuria and diabe-
tes at risk for progression, allowing the sorting of patients
that progressed toward overt DN over a 3-year period, and
(2) allowed the differentiation of DN and other chronic
renal diseases with 81% sensitivity and 91% specificity,
thereby more closely mimicking the actual clinical
situation where only rarely patients need to be distin-
guished from healthy controls. The data were subsequent-
ly confirmed in several independent studies ([49] and
Zürbig et al., in preparation). These CE–MS-selected
urinary biomarkers thus clearly have a potential for use
in the clinic and are also potentially applicable in the
pediatric population, as shown in a small pilot study [59].
Encouraged by these data, our group is now focusing on
32 Pediatr Nephrol (2010) 25:27–35

7. testing age-corrected adult biomarkers [51] of DN in a
type I diabetic pediatric cohort.
Additional studies in adult cohorts that resulted in
apparently valid biomarkers which may well be relevant
in the pediatric population have been carried out on
biomarkers for chronic kidney disease [53], immuno-
globulin (Ig)A-nephropathy [60, 61] and the detection of
acute rejection of kidney transplants [62].
Urinary biomarkers of diseases from extra renal sites
It has been estimated that approximately 30% of the
proteins and peptides in the urine originate from the
circulation. This has been exploited for the identification
of biomarkers of cardiovascular disease in adults. As
cardiovascular co-morbidity may concern the pediatric
population of children with early onset chronic kidney
disease (CKD), we would like to highlight an example of
the identification and independent validation of urinary
biomarkers for cardiovascular disease.
Coronary artery disease (CAD) is a leading cause of
morbidity and mortality worldwide. Despite multiple clinical,
electrographic and biochemical characteristics, there are
subgroups of patients who progress to severe, life-
threatening CAD without clinically overt symptoms and signs
[63]. For example, patients with type II diabetes and the
elderly frequently suffer from silent myocardial infarctions
with significantly increased risk of complications [64]. Early
diagnosis of CAD in its pre-symptomatic stage would allow
for better targeted and, therefore, more effective primary
prevention than what is possible with current clinical
recommendations. Urinary biomarkers for CAD have been
recently defined and validated in an independent population
[12]. In this study, urine from 88 CAD patients and 282
controls was examined by CE-MS, resulting in the identifi-
cation of 15 peptides that defined a characteristic CAD
signature panel. In a second step, this panel was evaluated in
a blinded study on 47 CAD patients and 12 healthy
individuals. The CAD patients were identified with 90%
sensitivity and specificity. In addition, the polypeptide CAD
signature panel significantly changed after therapeutic
intervention towards the polypeptide signature of healthy
humans. Recent data show that patients with CAD could be
distinguished from patients presenting symptoms of CAD
but without clinical evidence on the coronary angiography
[65]. The prospective value of the urinary proteomics for
CAD was further validated in prospectively collected
samples from patients with type I diabetes [49]. In this
blinded study, the data clearly show that urinary proteome
analysis can also provide useful biomarkers for diseases
more distant from the kidney and the urinary tract.
Outlook
Recent progress in mass spectrometry and biomarker
discovery has enabled the identification of urinary biomarkers
of (renal) disease that have the potential to be used in non-
invasive diagnostic and prognostic tests. The number of
published studies employing both separate discovery and
validation cohorts and using adapted statistics is, however,
still limited. Unfortunately, this type of research is as yet
underrepresented in the pediatric population where funding is
scarce. Non-invasive analyses are needed most urgently for
several reasons: (1) non-invasive detection will significantly
increase patient comfort and be highly appreciated by both
children and parents, (2) non-invasive procedures will
facilitate the close surveillance of individuals at risk and thus
identify patients at an early stage of disease progression,
thereby allowing individually tailored treatment or follow-up
of these individuals and (3) early non-invasive detection is
expected to reduce the costs of medical care.
However, since the currently available data clearly
demonstrate the potential of urinary proteomics, especially
in the pediatric population, the technologies are sufficiently
advanced to apply them with a good chance for success. As
the need for such biomarkers is undisputed, we anticipate
that valid reports on urinary biomarkers for several
pediatric diseases will be published in the near future. It is
to be hoped that such efforts, which are likely to succeed,
will find support from funding agencies, even though they
target only a minor fraction of the general population.
Acknowledgments CC, CL, SD and JPS acknowledge financial
support from the Agence Nationale pour la Recherche (ANR-07-
PHYSIO-004-01), the Fondation pour la Recherche Médicale “Grands
Equipements pour la Recherche Biomédicale” and the CPER2007–2013
programme. The work of SD was sponsored by the Inserm Interface
program. HM was supported in part by EUROTRANS-BIO grant ETB-
2006-016 and EU Funding through InGenious HyperCare (LSHM-C7-
2006-037093) and PREDICTIONS (1272568). JPS was supported by
Inserm, the “Direction Régional Clinique” (CHU de Toulouse, France)
under the Interface program and by the Fondation pour la Recherche
Médicale.
Conflict of interest statement Harald Mischak is the co-founder
and co-owner of Mosaiques Diagnostics, who developed the CE-MS
technology for clinical applications.
References
1. Petricoin EF, Ardekani AM, Hitt BA, Levine PJ, Fusaro VA,
Steinberg SM, Mills GB, Simone C, Fishman DA, Kohn EC,
Liotta LA (2002) Use of proteomic patterns in serum to identify
ovarian cancer. Lancet 359:572–577
2. Check E (2004) Proteomics and cancer: running before we can
walk? Nature 429:496–497
3. Villanueva J, Shaffer DR, Philip J, Chaparro CA, Erdjument-
Bromage H, Olshen AB, Fleisher M, Lilja H, Brogi E, Boyd J,
Pediatr Nephrol (2010) 25:27–35 33

8. Sanchez-Carbayo M, Holland EC, Cordon-Cardo C, Scher HI,
Tempst P (2006) Differential exoprotease activities confer tumor-
specific serum peptidome patterns. J Clin Invest 116:271–284
4. Kolch W, Neususs C, Pelzing M, Mischak H (2005) Capillary
electrophoresis-mass spectrometry as a powerful tool in clinical
diagnosis and biomarker discovery. Mass Spectrom Rev 24:
959–977
5. Omenn GS, States DJ, Adamski M, Blackwell TW, Menon R,
Hermjakob H, Apweiler R, Haab BB, Simpson RJ, Eddes JS,
Kapp EA, Moritz RL, Chan DW, Rai AJ, Admon A, Aebersold R,
Eng J, Hancock WS, Hefta SA, Meyer H, Paik YK, Yoo JS, Ping P,
Pounds J, Adkins J, Qian X, Wang R, Wasinger V, Wu CY, Zhao X,
Zeng R, Archakov A, Tsugita A, Beer I, Pandey A, Pisano M,
Andrews P, Tammen H, Speicher DW, Hanash SM (2005)
Overview of the HUPO Plasma Proteome Project: results from the
pilot phase with 35 collaborating laboratories and multiple
analytical groups, generating a core dataset of 3020 proteins and a
publicly-available database. Proteomics 5:3226–3245
6. Righetti PG, Boschetti E, Lomas L, Citterio A (2006) Protein
Equalizer Technology: the quest for a “democratic proteome”.
Proteomics 6:3980–3992
7. Schaub S, Wilkins J, Weiler T, Sangster K, Rush D, Nickerson P
(2004) Urine protein profiling with surface-enhanced laser-
desorption/ionization time-of-flight mass spectrometry. Kidney
Int 65:323–332
8. Theodorescu D, Wittke S, Ross MM, Walden M, Conaway M,
Just I, Mischak H, Frierson HF (2006) Discovery and validation
of new protein biomarkers for urothelial cancer: a prospective
analysis. Lancet Oncol 7:230–240
9. Vestergaard P, Leverett R (1958) Constancy of urinary creatinine
excretion. J Lab Clin Med 51:211–218
10. Schiffer E, Mischak H, Novak J (2006) High resolution proteome/
peptidome analysis of body fluids by capillary electrophoresis
coupled with MS. Proteomics 6:5615–5627
11. Thongboonkerd V, Malasit P (2005) Renal and urinary proteomics:
current applications and challenges. Proteomics 5:1033–1042
12. Zimmerli LU, Schiffer E, Zurbig P, Good DM, Kellmann M,
Mouls L, Pitt AR, Coon JJ, Schmieder RE, Peter K, Mischak H,
Kolch W, Delles C, Dominiczak AF (2007) Urinary proteomic
biomarkers in coronary artery disease. Mol Cell Proteomics
7:290–298
13. Decramer S, Gonzalez de Peredo A, Breuil B, Mischak H,
Monsarrat B, Bascands JL, Schanstra JP (2008) Urine in clinical
proteomics. Mol Cell Proteomics 7:1850–1862
14. Thongboonkerd V (2007) Recent progress in urinary proteomics.
Proteomics Clin Appl 1:780–791
15. Good DM, Thongboonkerd V, Novak J, Bascands JL, Schanstra JP,
Coon JJ, Dominiczak A, Mischak H (2007) Body fluid proteomics
for biomarker discovery: lessons from the past hold the key to
success in the future. J Proteome Res 6:4549–4555
16. Mischak H, Apweiler R, Banks RE, Conaway M, Coon JJ,
Dominizak A, Ehrich JH, Fliser D, Girolami M, Hermjakob H,
Hochstrasser DF, Jankowski V, Julian BA, Kolch W, Massy Z,
Neususs C, Novak J, Peter K, Rossing K, Schanstra JP, Semmes
OJ, Theodorescu D, Thongboonkerd V, Weissinger EM, Van Eyk
JE, Yamamoto T (2007) Clinical proteomics: a need to define the
field and to begin to set adequate standards. Proteomics Clin Appl
1:148–156
17. Thongboonkerd V (2007) Practical points in urinary proteomics. J
Proteome Res 6:3881–3890
18. Dakna M, He Z, Yu WC, Mischak H, Kolch W (2009) Technical,
bioinformatical and statistical aspects of liquid chromatography-
mass spectrometry (LC-MS) and capillary electrophoresis-mass
spectrometry (CE-MS) based clinical proteomics: a critical
assessment. J Chromatogr B Anal Technol Biomed Life Sci
877:1250–1258
19. Mischak H, Coon JJ, Novak J, Weissinger EM, Schanstra JP,
Dominiczak AF (2009) Capillary electrophoresis-mass spectrom-
etry as a powerful tool in biomarker discovery and clinical
diagnosis: an update of recent developments. Mass Spectrom Rev.
doi:10.1002/mas.20205
20. Hernandez J, Thompson IM (2004) Prostate-specific antigen: a
review of the validation of the most commonly used cancer
biomarker. Cancer 101:894–904
21. Caramori ML, Fioretto P, Mauer M (2000) The need for early
predictors of diabetic nephropathy risk: is albumin excretion rate
sufficient? Diabetes 49:1399–1408
22. Mogensen CE (1994) Systemic blood pressure and glomerular
leakage with particular reference to diabetes and hypertension. J
Intern Med 235:297–316
23. Rossing K (2007) Progression and remission of nephropathy in
type 2 diabetes: new strategies of treatment and monitoring. Dan
Med Bull 54:79–98
24. Fliser D, Novak J, Thongboonkerd V, Argiles A, Jankowski V,
Girolami MA, Jankowski J, Mischak H (2007) Advances in
urinary proteome analysis and biomarker discovery. J Am Soc
Nephrol 18:1057–1071
25. Malik G, Ward MD, Gupta SK, Trosset MW, Grizzle WE, Adam
BL, Diaz JI, Semmes OJ (2005) Serum levels of an isoform of
apolipoprotein A-II as a potential marker for prostate cancer. Clin
Cancer Res 11:1073–1085
26. Chang CP, McDill BW, Neilson JR, Joist HE, Epstein JA,
Crabtree GR, Chen F (2004) Calcineurin is required in urinary
tract mesenchyme for the development of the pyeloureteral
peristaltic machinery. J Clin Invest 113:1051–1058
27. Csaicsich D, Greenbaum LA, Aufricht C (2004) Upper urinary
tract: when is obstruction obstruction? Curr Opin Urol 14:213–
217
28. Chertin B, Pollack A, Koulikov D, Rabinowitz R, Hain D, Hadas-
Halpren I, Farkas A (2006) Conservative treatment of ureter-
opelvic junction obstruction in children with antenatal diagnosis
of hydronephrosis: lessons learned after 16 years of follow-up.
Eur Urol 49:734–738, discussion 739
29. Decramer S, Bascands JL, Schanstra JP (2007) Non-invasive
markers of ureteropelvic junction obstruction. World J Urol
25:457–465
30. Decramer S, Wittke S, Mischak H, Zurbig P, Walden M, Bouissou F,
Bascands JL, Schanstra JP (2006) Predicting the clinical outcome of
congenital unilateral ureteropelvic junction obstruction in newborn
by urinary proteome analysis. Nat Med 12:398–400
31. Maizels M, Reisman ME, Flom LS, Nelson J, Fernbach S, Firlit CF,
Conway JJ (1992) Grading nephroureteral dilatation detected in the
first year of life: correlation with obstruction. J Urol 148:609–614,
discussion 615–606
32. Sibai H, Salle JL, Houle AM, Lambert R (2001) Hydronephrosis
with diffuse or segmental cortical thinning: impact on renal
function. J Urol 165:2293–2295
33. Mesrobian HG (2008) The value of newborn urinary proteome
analysis in the evaluation and management of ureteropelvic
junction obstruction: a cost-effectiveness study. World J Urol
27:379–383
34. Van’t Hoff WG (2000) Molecular developments in renal tubulo-
pathies. Arch Dis Child 83:189–191
35. Lee S, Lee T, Lee B, Choi H, Yang M, Ihm CG, Kim M (2004)
Fanconi's syndrome and subsequent progressive renal failure
caused by a Chinese herb containing aristolochic acid. Nephrol-
ogy (Carlton) 9:126–129
36. Watanabe T, Yoshikawa H, Yamazaki S, Abe Y, Abe T (2005)
Secondary renal Fanconi syndrome caused by valproate therapy.
Pediatr Nephrol 20:814–817
37. Choudhury D, Ahmed Z (2006) Drug-associated renal dysfunction
and injury. Nat Clin Pract Nephrol 2:80–91
34 Pediatr Nephrol (2010) 25:27–35

9. 38. Izzedine H, Launay-Vacher V, Isnard-Bagnis C, Deray G (2003)
Drug-induced Fanconi’s syndrome. Am J Kidney Dis 41:292–309
39. Drube J, Schiffer E, Mischak H, Kemper MJ, Neuhaus T, Pape L,
Lichtinghagen R, Ehrich JH (2009) Urinary proteome pattern in
children with renal Fanconi syndrome. Nephrol Dial Transplant
24:2161–2169
40. Rossi R, Ehrich JH (1993) Partial and complete de Toni-Debre-
Fanconi syndrome after ifosfamide chemotherapy of childhood
malignancy. Eur J Clin Pharmacol 44[Suppl 1]:S43–S45
41. Skinner R, Cotterill SJ, Stevens MC (2000) Risk factors for
nephrotoxicity after ifosfamide treatment in children: a UKCCSG
Late Effects Group study. United Kingdom Children’s Cancer
Study Group. Br J Cancer 82:1636–1645
42. Caron HN, Abeling N, van Gennip A, de Kraker J, Voute PA
(1992) Hyperaminoaciduria identifies patients at risk of develop-
ing renal tubular toxicity associated with ifosfamide and platinate
containing regimens. Med Pediatr Oncol 20:42–47
43. Rossi R (1997) Nephrotoxicity of ifosfamide-moving towards
understanding the molecular mechanisms. Nephrol Dial Trans-
plant 12:1091–1092
44. Church DN, Hassan AB, Harper SJ, Wakeley CJ, Price CG (2007)
Osteomalacia as a late metabolic complication of Ifosfamide
chemotherapy in young adults: illustrative cases and review of the
literature. Sarcoma 2007:91586
45. Cutillas PR, Chalkley RJ, Hansen KC, Cramer R, Norden AG,
Waterfield MD, Burlingame AL, Unwin RJ (2004) The urinary
proteome in Fanconi syndrome implies specificity in the reab-
sorption of proteins by renal proximal tubule cells. Am J Physiol
Renal Physiol 287:F353–364
46. Vilasi A, Cutillas PR, Maher AD, Zirah SF, Capasso G, Norden AW,
Holmes E, Nicholson JK, Unwin RJ (2007) Combined proteomic
and metabonomic studies in three genetic forms of the renal Fanconi
syndrome. Am J Physiol Renal Physiol 293:F456–F467
47. Rossing K, Mischak H, Dakna M, Zurbig P, Novak J, Julian BA,
Good DM, Coon JJ, Tarnow L, Rossing P (2008) Urinary
proteomics in diabetes and CKD. J Am Soc Nephrol 19:1283–1290
48. Rossing K, Mischak H, Parving HH, Christensen PK, Walden M,
Hillmann M, Kaiser T (2005) Impact of diabetic nephropathy and
angiotensin II receptor blockade on urinary polypeptide patterns.
Kidney Int 68:193–205
49. Snell-Bergeon JK, Maahs DM, Ogden LG, Kinney GL, Hokanson JE,
Schiffer E, Rewers M, Mischak H (2009) Evaluation of urinary
biomarkers for coronary artery disease, diabetes, and diabetic kidney
disease. Diabetes Technol Ther 11:1–9
50. Merchant ML, Klein JB (2007) Proteomics and diabetic
nephropathy. Semin Nephrol 27:627–636
51. Zurbig P, Decramer S, Dakna M, Jantos J, Good DM, Coon JJ,
Bandin F, Mischak H, Bascands JL, Schanstra JP (2009) The
human urinary proteome reveals high similarity between kidney
aging and chronic kidney disease. Proteomics 9:2108–2117
52. Rodwell GE, Sonu R, Zahn JM, Lund J, Wilhelmy J, Wang L,
Xiao W, Mindrinos M, Crane E, Segal E, Myers BD, Brooks JD,
Davis RW, Higgins J, Owen AB, Kim SK (2004) A transcriptional
profile of aging in the human kidney. PLoS Biol 2:e427
53. Coon JJ, Zurbig P, Dakna M, Dominiczak AF, Decramer S, Fliser
D, Frommberger M, Golovko I, Good DM, Herget-Rosenthal S,
Jankowski J, Julian BA, Kellmann M, Kolch W, Massy Z, Novak
J, Rossing K, Schanstra JP, Schiffer E, Theodorescu D, Vanholder
R, Weissinger EM, Mischak H, Schmitt-Kopplin P (2008) CE-MS
analysis of the human urinary proteome for biomarker discovery
and disease diagnostics. Proteomics Clin Appl 2:964–973
54. Pinhas-Hamiel O, Zeitler P (2007) Acute and chronic complica-
tions of type 2 diabetes mellitus in children and adolescents.
Lancet 369:1823–1831
55. EURODIAB ACE Study Group (2000) Variation and trends in
incidence of childhood diabetes in Europe. EURODIAB ACE
Study Group. Lancet 355:873–876
56. Aanstoot HJ, Anderson BJ, Daneman D, Danne T, Donaghue K,
Kaufman F, Rea RR, Uchigata Y (2007) The global burden of
youth diabetes: perspectives and potential. Pediatr Diabetes 8
[Suppl 8]:1–44
57. Onkamo P, Vaananen S, Karvonen M, Tuomilehto J (1999)
Worldwide increase in incidence of Type I diabetes-the analysis of
the data on published incidence trends. Diabetologia 42:
1395–1403
58. Rossing P (2005) The changing epidemiology of diabetic micro-
angiopathy in type 1 diabetes. Diabetologia 48:1439–1444
59. Meier M, Kaiser T, Herrmann A, Knueppel S, Hillmann M,
Koester P, Danne T, Haller H, Fliser D, Mischak H (2005)
Identification of urinary protein pattern in type 1 diabetic
adolescents with early diabetic nephropathy by a novel combined
proteome analysis. J Diabetes Complications 19:223–232
60. Haubitz M, Wittke S, Weissinger EM, Walden M, Rupprecht HD,
Floege J, Haller H, Mischak H (2005) Urine protein patterns can
serve as diagnostic tools in patients with IgA nephropathy. Kidney
Int 67:2313–2320
61. Julian BA, Wittke S, Novak J, Good DM, Coon JJ, Kellmann M,
Zurbig P, Schiffer E, Haubitz M, Moldoveanu Z, Calcatera SM,
Wyatt RJ, Sykora J, Sladkova E, Hes O, Mischak H, McGuire BM
(2007) Electrophoretic methods for analysis of urinary polypep-
tides in IgA-associated renal diseases. Electrophoresis 28:
4469–4483
62. Wittke S, Haubitz M, Walden M, Rohde F, Schwarz A, Mengel M,
Mischak H, Haller H, Gwinner W (2005) Detection of acute
tubulointerstitial rejection by proteomic analysis of urinary samples
in renal transplant recipients. Am J Transplant 5:2479–2488
63. Fazzini PF, Prati PL, Rovelli F, Antoniucci D, Menghini F,
Seccareccia F, Menotti A (1993) Epidemiology of silent myocar-
dial ischemia in asymptomatic middle-aged men (the ECCIS
Project). Am J Cardiol 72:1383–1388
64. Scognamiglio R, Negut C, Ramondo A, Tiengo A, Avogaro A
(2006) Detection of coronary artery disease in asymptomatic
patients with type 2 diabetes mellitus. J Am Coll Cardiol 47:65–71
65. von Zur Muhlen C, Schiffer E, Zuerbig P, Kellmann M, Brasse M,
Meert N, Vanholder RC, Dominiczak AF, Chen YC, Mischak H,
Bode C, Peter K (2009) Evaluation of urine proteome pattern
analysis for its potential to reflect coronary artery atherosclerosis
in symptomatic patients. J Proteome Res 8:335–345
Pediatr Nephrol (2010) 25:27–35 35