Blomme et al. Glycomics in paediatric NAFLD
1
Serum protein N-glycosylation in paediatric non-alcoholic fatty liver
disease
Short running head: Glycomics in paediatric NAFLD
Bram Blomme1*, Emer Fitzpatrick2*, Alberto Quaglia3, Ruth De Bruyne4, Anil Dhawan2**, Hans Van
Vlierberghe1**
1
Department of Hepatology and Gastroenterology, Ghent University Hospital, Ghent, Belgium
2
Paediatric Liver, GI, and Nutrition Centre, King’s College London School of Medicine at King’s College
Hospital, London, UK
3
Institute of Liver Studies, King’s College London School of Medicine at King’s College Hospital,
London, UK
4
Department of Paediatric Gastroenterology, Hepatology, and Nutrition, University Hospital Ghent,
Belgium
*
Both authors contributed equally to this work
**
Co-last authorship
Corresponding author:
Bram Blomme, PhD
Department of Gastroenterology and Hepatology
Ghent University Hospital
De Pintelaan 185
B-9000 Ghent, Belgium
Blomme et al. Glycomics in paediatric NAFLD
Tel.: +32 9 332 5301
Fax: +32 9 332 4984
bram.blomme@ugent.be
Type of manuscript: Original article
Word count manuscript: 4309
Word count abstract: 242
Number of references: 26
Number of Tables: 3
Number of Figures: 2
Conflict of interest statement: The authors declare no conflict of interest
2
Blomme et al. Glycomics in paediatric NAFLD
3
Abstract
Objective: We have previously shown the potential of glycomics to distinguish patients with
steatosis from patients with non-alcoholic steatohepatitis (NASH) in an adult population. The pattern
of disease in paediatric patients is distinct from adults. The objective of this study was to
characterize the N-glycomic profile of children with varying degrees of non-alcoholic fatty liver
disease (NAFLD) and identify potential biomarker profiles of disease. Methods: Serum protein Nglycosylation patterns of 51 paediatric NAFLD patients were assessed with DNA-sequencer assisted
fluorophore-assisted capillary electrophoresis and compared with histology. Results: Peak 1 (NGA2F)
is the most significantly elevated N-glycan in paediatric NASH patients with peak 5 (NA2)
demonstrating the largest decrease. The logarithmically transformed ratio of peak 1 to peak 5 was 0.85 (SD 0.22) in patients with steatosis and borderline NASH and -0.73 (SD 0.12) in NASH (P=0.02).
The biomarker correlated well with the amount of lobular inflammation with a consistent increase of
marker score in ascending stage of lobular inflammation. There was also a trend in differentiating
patients with significant fibrosis ≥F2; -0.74 (SD 0.13) from patients with no/minimal fibrosis <F2; 0.86 (SD 0.24), P=0.06. Analysis of the N-glycans on immunoglobulin G confirmed the
undergalactosylation status typical for chronic inflammatory conditions. Conclusions: This study is
the first glycomic analysis performed in a paediatric NAFLD population. In agreement with the
results obtained in adults, B cells play a dominant role in the N-glycan alterations of paediatric NASH
patients.
Key words: Protein glycosylation; Immunoglobulin G; Glycomics; Electrophoresis, Capillary; Liver
steatosis; Paediatrics
Blomme et al. Glycomics in paediatric NAFLD
4
Introduction
In recent years, there has been an increased prevalence of paediatric obesity with a concomitant
increase of paediatric NAFLD. In fact, fatty liver is the most common liver abnormality in children
aged 2-19 years (1, 2). However, the true prevalence of paediatric NAFLD and NASH is unknown
given that the modalities used for diagnosis are not standardized (3). Liver biopsy is the generally
accepted standard for diagnosis and evaluation of severity but is clearly not a feasible tool for
frequent monitoring of disease nor for large scale screening. In most cases, aminotransferase
elevation in the absence of markers of any other liver disease, was used as a surrogate marker of
fatty liver disease (5, 6). Other markers, such as insulin resistance index, could have a role in
screening (7).
However, the spectrum of fatty liver encompass a wider range of subjects than identified by
elevated serum liver chemistries. Franzese et al demonstrated that in 53% of obese children
identified by ultrasound, only 32% had abnormalities in serum aminotransferases (8). Furthermore,
using magnetic resonance imaging, Burgert et al showed that only 48% of obese children with
intrahepatic fat accumulation had abnormal ALT levels (9). Moreover, NASH was diagnosed in 59% of
adult NAFLD patients with normal ALT levels indicating that routine liver function tests are not
reliable to identify NASH or more severe disease (10). Clearly, aminotransferase levels have very
limited sensitivity and specificity in the diagnosis of non-alcoholic fatty liver disease, and more
importantly, it can not distinguish the severity of disease. This is particularly important for long-term
monitoring and evaluation of therapeutic intervention. Therefore, the search for well-performing
non-invasive markers for paediatric NASH continues.
Interest in glycomics is growing steadily as a high throughput novel technology in the identification
of disease biomarkers (11, 12). Glycans are carbohydrate sequences that are conjugated to proteins
and lipids and are possible the most diverse class of molecule and glycomics is the study of this
diversity. Glycosylation is the post-translational modification of secreted proteins. Changes in
Blomme et al. Glycomics in paediatric NAFLD
5
glycosylation serve as a particularly good marker of liver dysfunction because most glycoprotein in
serum (aside from IgG) are made in the liver. Thus, the N-glycome profile will reflect any changes in
either the liver or B cell function. We previously performed a glycomic investigation in an adult
NAFLD population. A pilot study including bariatric surgery patients led to the development of a
glycomarker that could distinguish NASH from steatosis independently of fibrosis. This glycomarker
was the ratio of two N-glycans, NGA2F, a fully agalactosylated N-glycan that is exclusively present on
immunoglobulin G (IgG) and NA2, the most abundant N-glycan on liver-produced protein, but
present in low quantity on IgG. This biomarker was subsequently validated in a large, independent
clinical NAFLD population and multivariate analysis showed that our glycomarker was an
independent predictor of NASH (13).
A paediatric NAFLD population displays several differences with its adult counterpart, especially at
the histological level. Schwimmer et al examined the histological appearance of 100 paediatric cases
and identified two types of steatohepatitis (14). Type 1 NASH was consistent with NASH as described
in adults and was characterized by steatosis, ballooning degeneration and perisinusoidal fibrosis in
the absence of portal changes. In contrast, type 2 NASH was characterized by steatosis, portal
inflammation, and/or portal fibrosis in the absence of ballooning degeneration and perisinusoidal
fibrosis. Type 1 NASH was reported to be present in only 17% of paediatric NAFLD, whereas type 2
NASH was present in 51%. Type 1 and type 2 NASH may have a different pathogenesis, natural
history and response to treatment (14).
The aim of this study was to characterize the glycomic profile of children with varying degrees of
NAFLD and identify potential biomarker profiles of disease.
Blomme et al. Glycomics in paediatric NAFLD
6
Patients and Methods
Pediatric NAFLD population
Children with biopsy proven non-alcoholic fatty liver disease (n=51) were recruited from a tertiary
care paediatric hepatology unit. Children with suspected NAFLD underwent a Menghini liver biopsy
because of abnormal transaminases and/or splenomegaly. Biopsies were routinely processed;
paraffin embedded sections were stained with haematoxylin and eosin, reticulin, orcein and Perls
staining. In addition, fresh tissue (>1 mg) was sent for copper quantification to rule out Wilson
disease. A diagnosis of NAFLD was made on the basis of typical histological findings (15) by a
hepatohistopathologist in the appropriate setting and following exclusion of other liver disease.
None of the children drank alcohol. The following anthropomorphic data were collected from each
child: height, weight and BMI z-score. Biochemical data included HOMA-IR, renal, liver and bone
profiles, lipid profiles and full blood count. Finally, the echogenicity of the liver, spleen size and any
anatomical abnormalities were recorded using ultrasound. For this study, additional serum was
taken either on or within 30 days of biopsy. Serum was collected, centrifuged and stored at -80 °C
until analysis. Informed consent was given by the caregiver and the study was approved by the
Ethical Committee of the Ghent University Hospital and the National Research Ethics Committee
(UK) as part of a larger study on biomarkers of paediatric NAFLD.
Liver biopsy scoring
Histological specimens were scored according to NAFLD activity score (NAS), also scoring for stage of
fibrosis (16). This is the non-weighted sum of steatosis (0-3), ballooning (0-2), and lobular
inflammation (0-3). NASH is defined as a score ≥5, ’not NASH’ as a score ≤2, and a score of 3 or 4 is
classified as “borderline NASH”. This scoring system does not replace the histopathologist in making
a diagnosis of NASH, but rather is useful in reproducibly stratifying severity of disease activity.
Fibrosis was scored using a 5-point scale: 0, no fibrosis; 1, mild/moderate perisinusoidal or portal
Blomme et al. Glycomics in paediatric NAFLD
7
fibrosis; 2, both perisinusoidal and portal fibrosis; 3, bridging fibrosis and 4, cirrhosis. A single
histopathologist scored the specimens in batches of 30 to minimize intra-observer variability and
was blinded to other markers. Distinction between NASH type 1 and 2 was made as previously
described (14).
IgG depletion
Serum samples were depleted using beads covered with protein A/G (Thermo Scientific, Waltham,
MA, USA). 100 µl of serum was diluted with 100 µl of binding buffer and subsequently incubated
with 50 µl beads for 1 hour. The mixture was transferred to a 96-well filter plate and centrifuged at
1000g for 15 seconds. The IgG depleted eluate was captured in a microtiter plate. Subsequently,
after five wash steps with binding buffer, pure IgG was eluted from the beads after incubation (5
min) with 0.1M glycine pH 2 in the 96-well filter plate. After centrifugation at 1000g for 15 seconds,
the eluate is neutralized with 1M Tris pH 8.8.
Glycomic analysis
For an elaborate description of the protocol, we refer to Laroy et al (17). Briefly, the N-glycans
present in 5 µl of serum or pure IgG elution fraction were released from the proteins with peptide Nglycanase F. Subsequently, the N-glycans are fluorescently labeled and desialylated. The labeled
glycans were profiled and analyzed by DNA sequencer-assisted fluorophore-assisted capillary
electrophoresis (DSA-FACE) technology. In analogy with adult patients, we could observe thirteen
peaks in the total serum electropherogram and eight peaks in the IgG electropherogram. The height
of every peak was quantified to obtain a numerical description of the profiles and these were
normalized to the total intensity of the measured peaks (represented as a percentage of the total
peak height). Structural characterization of the fluorescently labeled N-glycans was done as
previously described (18).
Blomme et al. Glycomics in paediatric NAFLD
8
Statistical analysis
SPSS v17.0 was used for analysis. Descriptive results are expressed as median and interquartile range
or number (percentage) of patients with a condition. In case of the logarithmically transformed
Glycomarker, the mean and standard deviation are expressed as it followed a normal distribution.
The Mann Whitney U Test, Fischer’s exact test or the Kruskall Wallis test were used to compare non
parametric categorical / continuous data. Pearson’s chi square test, Fischer’s exact test or one-way
ANOVA were used to compare parametric data. Liver biopsy was used as the criterion standard. A P
value of <0.05 was considered significant.
Results
Demographics and clinical characteristics
Fifty-one children were recruited for this study with a median age of 13.3 years (range 4.5 – 17.4).
Thirty-one (61%) were boys. The median BMI z-score of the group was 1.81. Seventy-five percent of
the paediatric population had splenomegaly at time of biopsy. No child had abnormal liver synthetic
function, evidence of varices or decompensated liver disease. Eighty three percent of children had
evidence of abnormal insulin sensitivity (HOMA-IR>3). The overall median HOMA-IR was 4.83. Fortyseven pediatric patients had fibrosis (92.2%). Thirty-five of these patients were diagnosed with type
2 NASH (74.5%), 11 had mixed features (23.4%) and only 1 patient had proper type 1 NASH (2.1%).
All clinical and demographic data are summarized in Table 1.
Histological analysis
Five children scored as ‘not NASH’ or simple steatosis according to the Kleiner scoring system, 18 as
borderline NASH and 28 as true NASH. To compensate for low numbers in the simple steatosis
group, this was merged with the borderline NASH group in order to reflect relative severity of
Blomme et al. Glycomics in paediatric NAFLD
9
disease. Ten children (19.6%) scored 0 for inflammation, 29 (56.9%) scored 1 (mild inflammation)
and 12 (23.5%) scored 2 (moderate inflammation). No child had severe inflammation on biopsy. Four
children did not have evidence of fibrosis, fourteen scored fibrosis stage 1, thirteen as fibrosis stage
2, nineteen as fibrosis stage 3 and one child as cirrhotic. Children were classified as having no /
minimal fibrosis (<F2), n=18, or as significant fibrosis (≤ F2), n=33. Again, this was to preserve clinical
relevance in the study. Histological data are summarized in Table 2.
Total serum glycomic analysis in NASH versus borderline NASH and steatosis
Analysis of the total serum N-glycome revealed two agalactosylated glycans, peak 1 (NGA2F) and
peak 4 (NGA1A2F), that were significantly increased in abundance in children with NASH versus the
remainder (P=0.011 and P=0.033, respectively). Peak 9´ (NA3Fc) was also significantly elevated in
this group (P=0.022) (Figure 1, Table 3).
IgG glycomic analysis in NASH versus borderline NASH and steatosis
N-glycomic analysis of the glycans present on IgG revealed that peak 1 (NGA2F) was similarly
significantly increased in NASH patients (P=0.024). Peak 6 (NA2F) was significantly decreased in
abundance in children with NASH versus the remainder (P=0.01) (Figure 1, Table 3).
Evaluation of serum Glycomarker test in paediatric population
Peak 1 (NGA2F) was the most significantly increased N-glycan in paediatric NASH patients with peak
5 (NA2) demonstrating the largest decrease (Table 3). This biomarker was subsequently evaluated in
the different patient groups. Patients with simple steatosis were found to have a mean score of -0.9
(±0.1), patients with borderline NASH had a mean score of -0.83 (±0.25) and NASH patients had a
mean score of -0.73 (±0.12). When the scores of patients with simple steatosis and borderline NASH
were merged, this group displayed a mean value of -0.85 (±0.22) which was significantly lower in
comparison with the NASH group (-0.73) (P=0.02). The area under the receiver operating curve
Blomme et al. Glycomics in paediatric NAFLD
10
(AUROC) to distinguish children with NASH versus the remainder was 0.72 (95% CI 0.57 – 0.86) and
0.79 (95% CI 0.64-0.94) to distinguish the group with simple steatosis (figure 2).
The Glycomarker also correlated well with the amount of inflammation in biopsies. Mean score in
patients with no inflammation was -0.88 (±0.27), mild inflammation was -0.77 (±0.17) and moderate
inflammation was -0.74 (±0.08). There was also an increase in marker score in paediatric patients
with ballooning versus no ballooning. Mean score in patients with no ballooning was -0.83 (±0.1),
with few balloon cells was -0.78 (±0.18) and with prominent ballooning was -0.78 (±0.2).
There was a trend towards significance in differentiating the group with significant fibrosis (≥F2 –
n=18); -0.74 (±0.13) from the group with no/minimal fibrosis (<F2 – n=33); -0.86 (±0.24) (P=0.06).
Discussion
Non-invasive methods to screen an obese, paediatric population for NAFLD are not readily available.
Most research efforts have focused on adult NAFLD, however the paediatric population is not far
behind in terms of prevalence with up to 10% of the children affected (2, 19). Moreover, the
problems of adult NAFLD patients generally originate in childhood, mostly during adolescence, and a
correction in lifestyle or even clinical treatment in this early phase would be beneficial for the
management of the disease. This study sets out to distinguish the differing degree of disease
severity in children with NAFLD. This is important as the prognosis for those with simple steatosis
may not be much different from the general population but is significantly worse in terms of both all
cause and liver related mortality for those with NASH (20). The severity of the condition will change
over time and noninvasive biomarkers must be effective in monitoring disease progression /
regression.
The NAS scoring system, while designed primarily to reflect changes over time in the context of
clinical trials, is the best available tool against which to compare a biomarker. However, we
Blomme et al. Glycomics in paediatric NAFLD
11
emphasize that neither the diagnosis of NAFLD nor the presence of steatohepatitis can not be
inferred from the NAS and requires an overall assessment of the presence and distribution of the
individual histological findings (21). Despite this, not one scoring system has been properly validated
and the NAS is the most widely used in studies which evaluate biomarkers in paediatric and adult
NAFLD patients (22-26). This makes comparison of diagnostic efficacy between biomarkers feasible.
Our data indicate that paediatric NAFLD patients show very similar results in a glycomic analysis
compared to adult NAFLD patients with an increase of agalactosylated glycans in NASH patients. In
this pediatric cohort, two of these N-glycans (NGA2F and NGA1A2F) reached significance. This
undergalactosylation is present on IgG and its glycomic profile confirmed the increase of NGA2F in
favor of the decrease of NA2F (figure 1). This is similar to the pattern found in other chronic
inflammatory diseases such as rheumatoid arthritis, ankylosing spondylitis and Crohn’s disease (2730). This reflects the inflammatory component to the disease and explains the association with
steatohepatitis rather than with fibrosis. We also found a clear trend in differentiating patients with
no/minimal fibrosis from patients with significant fibrosis. However, evaluation of the distribution of
the different categories of lobular inflammation showed that nearly all patients with significant
fibrosis had lobular inflammation, while more than a quarter of the patients with no/minimal fibrosis
had no lobular inflammation.
An important issue in this study was the different pattern of fibrosis in children, which was type 2 in
75% and mixed in 21%. The similarity of the glycomic analysis in adult and paediatric patients despite
the clearly different liver histology was an extra confirmation that IgG N-glycosylation predominantly
determines the N-glycan alterations of NASH patients, whereas the N-glycosylation of liver-produced
protein was only marginally affected. A larger multi-centre glycomic study of children with NAFLD
may allow the development of biomarkers specific to the type 2 pattern of injury.
In agreement with the results in adult NAFLD patients, the log(peak1/peak5) biomarker displayed
the best result in distinguishing NASH from steatosis. The AUC to differentiate steatosis from
Blomme et al. Glycomics in paediatric NAFLD
12
borderline NASH and NASH was high at 0.79. However, the analysis to distinguish steatosis and
borderline NASH patients from NASH patients will provide a better representation of real-life
practice because of the small number of steatosis patients. This analysis still had a reasonable AUC
of 0.72.
Our glycomarker showed a good correlation with the amount of inflammation with a consistent,
gradual increase of mean marker score in ascending amount of lobular inflammation. This
observation is important because the appearance of an inflammatory reaction defines the onset of
NASH (31). Importantly in this context, the most elaborate difference was observed between
patients with no inflammation and patients with a mild inflammation, mean marker score does not
augment further when patients progressed to a moderate amount of lobular inflammation. This
suggests that the Glycomarker would be a good tool to early identify patients with simple steatosis
that progress to a more severe phenotype. The same was observed for the analysis in the different
stages of ballooning, although the difference between patients with or without ballooning was not
as extensive as in patients with or without inflammation.
The sample size is relatively small, however reflects the practice of a large tertiary paediatric liver
centre. The strengths of the study include the fact that these children all had biopsy-proven disease
in contrast to other studies in the field where the diagnosis of NAFLD was made on the basis of a
bright liver on ultrasound. The findings of this study will ideally be validated in a larger cohort.
In conclusion, the results in this study are novel in that they represent the first glycomic analysis of
paediatric NAFLD. They validate findings in adults with the condition in that a glycomarker can serve
as a biomarker of severity of disease in NAFLD. The strong involvement of B cells in the glycomic
alterations of patients with NASH supports previous research which indicates that NASH is a
systemic disease linked with visceral obesity and insulin resistance, rather than a liver-specific
disorder. Finally, this technology has recently been brought to a clinical platform in order that clinical
Blomme et al. Glycomics in paediatric NAFLD
13
chemistry laboratories will be able to use glycomics as a tool in the diagnosis and follow-up of NAFLD
patients (32).
Acknowledgements
BB receives a scholarship GOA BOFF07/GOA/017 of the University Ghent Research Fund (BOF).
References
1. Feldstein A, Charatcharoenwitthaya P, Treeprasertsuk S, Benson J, Enders F, Angulo P. The natural
history of non-alcoholic fatty liver disease in children: a follow-up study for up to 20 years. Gut.
2009; 58:1538-1544.
2. Patton H, Sirlin C, Behling C, Middleton M, Schwimmer J, Lavine J. Pediatric nonalcoholic fatty liver
disease: a critical appraisal of current data and implications for future research. J Pediatr
Gastroenterol Nutr. 2006; 43:413-427.
3. Oh M, Winn J, Poordad F. Review article: diagnosis and treatment of non-alcoholic fatty liver
disease. Aliment Pharmacol Ther. 2008; 28:503-522.
4. Schwimmer J. Definitive diagnosis and assessment of risk for nonalcoholic fatty liver disease in
children and adolescents. Semin Liver Dis. 2007; 27:312-318.
5. Schwimmer J, McGreal N, Deutsch R, Finigold M, Lavine J. Influence of gender, race, and ethnicity
on suspected fatty liver in obese adolescents. Pediatrics. 2005; 115:e561-e565.
6. Nobili V, Reale A, Alisi A, et al. Elevated serum ALT in children presenting to the emergency unit:
relationship with NAFLD. Dig Liver Dis. 2009; 41:749-752.
7. Reinehr T, Toschke A. Onset of puberty and cardiovascular risk factors in untreated obese children
and adolescents: a 1-year follow-up study. Arch Pediatr Adolesc Med. 2009; 163:709-715.
Blomme et al. Glycomics in paediatric NAFLD
14
8. Franzese A, Vajro P, Argenziano A, et al. Liver involvement in obese children. Ultrasonography and
liver enzyme levels at diagnosis and during follow-up in an Italian population. Dig Dis Sci. 1997;
42:1428-1432.
9. Burgert T, Taksaki S, Dziura J, et al. Alanine aminotransferase levels and fatty liver in childhood
obesity: associations with insulin resistance, adiponectin, and visceral fat. J Clin Endocrinol Metab.
2006; 91:4287-4294.
10. Fracanzani A, Valenti L, Bugianesi E, et al. Risk of severe liver disease in nonalcoholic fatty liver
disease with normal aminotransferase levels: a role for insulin and diabetes. Hepatology. 2008;
48:792-8.
11. Kam R, Poon T. The potentials of Glycomics in Biomarker Discovery. Clin Proteom. 2008; 4:67-79.
12. Blomme B, Vansteenkiste C, Callewaert N, Van Vlierberghe H. Alteration of protein glycosylation
in liver diseases. J Hepatol. 2009; 50:592-603.
13. Blomme B, Francque S, Trépo E, et al. N-glycan based biomarker distinguishing non-alcoholic
steatohepatitis from steatosis independently of fibrosis. Dig Liv Dis. In Press.
14. Schwimmer J, Behling C, Newburg R, et al. Histopathology of pediatric nonalcoholic fatty liver
disease. Hepatology. 2005; 42:641-649.
15. Brunt E. Pathology of nonalcoholic fatty liver disease. Nat Rev Gastroenterol Hepatol. 2010;
7(4):195-203.
16. Kleiner D, Brunt E, Van Natta M, et al. Design and validation of a histological scoring system of
nonalcoholic fatty liver disease. Hepatology. 2005; 41:1313-1321.
17. Laroy W, Contreras R, Callewaert N. Glycome mapping on DNA sequencing equipment. Nat
Protoc. 2006; 1:397-405.
Blomme et al. Glycomics in paediatric NAFLD
15
18. Callewaert N, Van Vlierberghe H, Van Hecke A, Laroy W, Delanghe J, Contreras R. Noninvasive
diagnosis of liver cirrhosis using DNA sequencer-based total serum protein glycomics. Nat Med.
2004; 10:429-434.
19. Schwimmer J, Deutsch R, Kahen T, Lavien J, Stanley C, Behling C. Prevalence of fatty liver in
children and adolescents. Pediatrics. 2006; 118:1388-1393.
20. Musso G, Gambino R, Cassadu M, Pagano G. A novel approach to control hyperglycemia in type 2
diabetes: Sodium glucose co-transport (SGLT) inhibitors. Systematic review and meta-analysis of
randomized trials. Ann Med. 2011 [Epub ahead of print].
21. Brunt E, Kleiner E, Wilson L, et al. Nonalcoholic fatty liver disease (NAFLD) activity score and the
histopathologic diagnosis in NAFLD: distinct clinicopathologic meanings. Hepatology 2011;53:810820.
22. Nobili V, Reale A, Alisi A, et al. Elevated serum ALT in children presenting to the emergency unit:
Relationship with NAFLD. Dig Liv Dis 2009;41:749-752.
23. Nobili V, Parola M, Alisi A, et al. Oxidative stress parameters in paediatric non-alcoholic fatty liver
disease. Int J Mol Med 2010;26:471-476.
24. Manousou P, Kalambokis G, Grilo F, et al. Serum ferritin is a discriminant marker for both fibrosis
and inflammation in histologically proven non-alcoholic fatty liver disease patients. Liver Int
2011;31:730-739.
25. Poynard T, Ratziu V, Charlotte F, et al. Diagnostic value of biochemical markers (NashTest) for the
prediction of non alcoholo steato hepatitis in patients with non-alcoholic fatty liver disease. BMC
Gastroenterol 2006;6:34-49.
26. Lirussi F, Vitturi N, Azzalini L, et al. Perihepatic adipose tissue thickness: a new non-invasive
marker of NAFLD. J Gastrointestin Liver Dis 2009;18:61-66.
Blomme et al. Glycomics in paediatric NAFLD
16
27. Van Beneden K, Coppieters K, Laroy W, et al. Reversible changes in serum immunoglobulin
galactosylation during the immune response and treatment of inflammatory autoimmune arthritis.
Ann Rheum Dis. 2008; 68:1360-1365.
28. Parekh R, Dwek R, Sutton B, et al. Association of rheumatoid arthritis and primary osteoarthritis
with changes in the glycosylation pattern of total serum IgG. Nature. 1985; 316:452-7.
29. Leirisalo-Repo M, Hernandez-Munoz HE, Rook GA. Agalactosyl IgG is elevated in patients with
active spondyloarthropathy. Rheumatol Int. 1999; 18:171-6.
30. Tomana M, Schrohenlohrer K, Koopman W, et al. Abnormal glycosylation of serum IgG from
patients with chronic inflammatory diseases. Arthritis Rheum. 1988; 31:333-8.
31. Tilg H, Moscher A. Evolution of inflammation in nonalcoholic fatty liver disease: the multiple
parallel hits hypothesis. Hepatology. 2010; 52:1836-1846.
32. Vanderschaeghe D, Szekrényes A, Wenz C, et al. High-throughput profiling of the serum Nglycome on capillary electrophoresis microfluids systems: towards clinical implementation of
GlycoHepatoTest. Anal Chem. 2010; 82:7408-7415.
Blomme et al. Glycomics in paediatric NAFLD
17
Figure legends
Figure 1. Representative total serum (A) and IgG (B) electropherogram of a paediatric patient with
simple steatosis and NASH. Peak 1 is an agalacto, core-α-1,6-fucosylated biantennary (NGA2F), peak
2 is an agalacto, core-α1,6-fucosylated bisecting biantennary (NGA2FB), peak 3 and peak 4 are single
agalacto, core-α-1,6-fucosylated biantennaries (NG1A2F), peak 5 is a bigalacto, biantennary (NA2),
peak 6 is a bigalacto, core-α-1,6-fucosylated biantennary (NA2F), peak 7 is a bigalacto, core-α-1,6fucosylated bisecting biantennary (NA2FB), peak 8 is a triantennary (NA3), peak 9 is a branching
α1,3-fucosylated triantennary (NA3Fb), peak 9’ is a core-α-1,6-fucosylated triantennary (NA3Fc),
peak 10 is branching α1,3-fucosylated and core α-1,6-fucosylated triantennary (NA3Fbc), peak 11 is a
tetra-antennary (NA4) and peak 12 is branching α1,3-fucosylated tetra-antennary (NA4Fb). The
symbols used in the structural formulas are: square indicates β-linked N-acetylglucosamine (GlcNAc);
yellow circle indicates β-linked galactose, triangle indicates α/β-1,3/6-linked fucose; green circle
indicates α/β-linked mannose.
Figure 2. Classification efficiency. ROC curve analysis to evaluate efficiency of Glycomarker in
differentiating between the group with steatosis and borderline NASH and the group with NASH
(Top) and in differentiating between the group with steatosis and the group with borderline NASH
and NASH (Bottom).
Blomme et al. Glycomics in paediatric NAFLD
18
Table 1: clinical and demographic characteristics of the paediatric NAFLD population
All (n=51)
Simple steatosis/
borderline NASH (n=23)
NASH (n=28)
P value
Age (y)
13.3 (11.6, 14.6)
13.7 (12.2, 14.6)
13.0 (11.5, 14.4)
0.416
Sex: male %
31 (60.8%)
14 (60.9%)
17 (60.7%)
0.991
BMI z-score
1.81 (1.44, 2.16)
1.84 (1.22, 2.2)
1.81 (1.53, 2.11)
0.656
Albumin (mg/ml)
48 (45, 49)
48 (44, 49.5)
47.5 (45.5, 49)
0.500
ALP (U/l)
274 (233, 319)
272 (226, 292)
299 (243, 349)
0.136
HOMA-IR
4.83 (3.59, 7.16)
4.9 (3.67, 6.95)
4.74 (3.53, 8.52)
0.891
Hemoglobin (g/dl)
13.4 (12.9, 14.2)
13.4 (12.9, 14.6)
13.4 (12.9, 13.8)
0.378
White blood count (x103/µl)
7.07 (6, 8.6)
6.83 (6, 8.8)
7.1 (5.98, 8.25)
0.691
Platelets (x103/µl)
323 (262, 368)
326 (263, 368)
310 (246, 379)
0.698
INR
1.0 (0.96, 1.03)
0.98 (0.92, 1.03)
1.0 (0.97, 1.03)
0.208
Tot Bil (µmol/l)
8 (6, 11)
9 (6, 11)
8 (6, 10)
0.71
AST (U/l)
58 (42, 84)
49 (36, 73)
61 (50, 105)
0.048
ALT (U/l)
73 (47, 103)
67 (38, 108)
76 (62, 101)
0.525
GGT (U/l)
36 (25, 64)
27 (18, 50)
41 (31, 76)
0.018
Cholesterol (mmol/l)
4.5 (4.0, 5.1)
4.9 (4.2, 5.2)
4.2 (3.7, 4.8)
0.021
Triglycerides (mmol/l)
1.7 (1.1, 2.7)
1.8 (1.35, 2.8)
1.65 (1.1, 2.7)
0.258
Splenomegaly (%)
38 (75%)
20 (87%)
18 (64%)
0.577
Glycomarker
-0.79 (0.18)
-0.85 (0.22)
-0.73 (0.12)
0.02
Blomme et al. Glycomics in paediatric NAFLD
19
Table 2: Histological findings in the paediatric study cohort (n=51)
All (n=51)
Simple steatosis/
NASH (n=28)
borderline NASH (n=23)
Steatosis
0
1
2
3
2 (3.9%)
15 (29.4%)
7 (13.7%)
27 (53.0%)
2 (8.7%)
15 (65.2%)
4 (17.4%)
2 (8.7%)
0 (0%)
0 (0%)
3 (10.7%)
25 (89.3%)
Inflammation
0
1
2
3
10 (19.6%)
29 (56.9%)
12 (23.5%)
0 (0%)
10 (43.5%)
12 (52.2%)
1 (4.3%)
0 (0%)
0 (0%)
17 (60.7%)
11 (39.3%)
0 (0%)
Ballooning
0
1
2
5 (9.8%)
20 (39.2%)
26 (51.0%)
5 (21.8%)
9 (39.1%)
9 (39.1%)
0 (0%)
11 (39.3%)
17 (60.7%)
Fibrosis
0
1
2
3
4
4 (7.8%)
14 (27.5%)
13 (25.5%)
19 (37.2%)
1 (2.0%)
2 (8.7%)
11 (47.8%)
4 (17.4%)
6 (26.1%)
0 (0%)
2 (7.2%)
3 (10.7%)
9 (32.1%)
13 (46.4%)
1 (3.6%)
Blomme et al. Glycomics in paediatric NAFLD
20
Table 3: Relative percentages of the different peaks in the total serum and IgG electropherogram
Total serum
Simple steatosis/
NASH (n=28)
Immunoglobulin G
P-value
borderline NASH (n=23)
Simple steatosis/
NASH (n=28)
P value
borderline NASH (n=23)
Peak 1 (%)
NGA2F
6.44 (5.65, 7.92)
8.25 (6.68, 9.06)
0.011
21.46 (17.02, 24.5)
23.41 (21.16, 26.71)
0.024
Peak 2 (%)
NGA2FB
1.15 (0.98, 1.41)
1.13 (1.01, 1.4)
0.88
2.75 (2.3, 3.46)
3.22 (2.75, 3.89)
0.14
Peak 3 (%)
NG1A2F
7.82 (6.76, 9.05)
7.62 (6.5, 9.08)
0.917
22.47 (20.7, 23.73)
22.93 (21.12, 24.18)
0.325
Peak 4 (%)
NGA1A2F
2.75 (2.3, 3.46)
3.4 (2.7, 3.84)
0.033
10.11 (8.6, 10.77)
10.45 (9.56, 11.57)
0.05
Peak 5 (%)
NA2
45.8 (40.88, 47.25)
43.77 (40.2, 45.5)
0.14
5.53 (4.91, 6.56)
5.46 (4.56, 6.21)
0.59
Peak 6 (%)
NA2F
20.93 (18.97, 24.38)
20.86 (19.26, 22.61)
0.798
32.79 (27.32, 35.14)
26.33 (25.27, 29.49)
0.01
Peak 7 (%)
NA2FB
4.17 (3.42, 4.85)
4.29 (3.58, 4.72)
0.545
5.61 (4.43, 7.06)
5.61 (4.68, 6.43)
0.776
Peak 8 (%)
NA3
6.85 (5.16, 7.53)
6.45 (5.12, 7.34)
0.806
0.28 (0.17, 0.46)
0.32 (0.18, 0.61)
0.466
Peak 9 (%)
NA3Fb
2.13 (1.34, 2.39)
1.64 (1.25, 2.39)
0.32
Peak 9´ (%)
NA3Fc
0.92 (0.72, 1.13)
1.37 (1.07, 1.68)
0.022
Peak 10 (%)
NA3Fbc
0.42 (0.32, 0.52)
0.51 (0.35, 0.67)
0.248
Peak 11 (%)
NA4
0.79 (0.68, 0.9)
0.86 (0.73, 0.99)
0.32
Peak 12 (%)
NA4F
0.4 (0.33, 0.48)
0.41 (0.34, 0.49)
0.992