Genetics, Immunology and biomarkers in clinical practice Peter Laszlo Lakatos 1st Department of Medicine Semmelweis University Budapest Hungary
Genetics, Immunology andbiomarkers in clinical practice
Peter Laszlo Lakatos1st Department of Medicine
Semmelweis University BudapestHungary
Why and where during the course of IBD are markers needed?
•Diagnosis and differential diagnosis?
•Prognosis and risk for complications?
•Assessement of disease activity?
•Optimazing drug therapy and side effects?
•Risk for post-operative recurrence?
Genetics
114 IBD patients (82, 71.9% with CD)-76.8% (86 of 112) would undergo testing for diagnostic confirmation-81.3% (91 of 112) for prognostic value-88.4% (99 of 112) for therapeutic decision making-85.0% (96 of 113) for advancement of medical knowledge- patients with a first-degree relative with IBD (raw score on self-willingness 4.62 vs 4.36; P = 0.026)
Konda V Inflamm Bowel Dis 2006;12:497-502.
Do patients want to be tested?
Classification of IBD…fiction?
IBD
Crohn’s
diease
Ulcerativecolitis
UC-like CD
IC
IBD… a complex phenotype
DizygoticMonozygotic
3.8%44.4%80Tysk et al
0%50%102Orholm et al
6.5%20%143Thompson et al
Concordance for diseaseN twinsReference
How to suspect genetic component?
Locus1(causal allele or genotype)
Monogenic versus Polygenic disease
Phenotype Phenotype
Locus2 Locus3 Locus3
Type II DM
CAPN10 HLA Susceptibility
Strong relationshipNo factor necessary/sufficient
to cause disease, allincrease individual’s risk
Effects of human functional genetic variation
Susceptibility… …neutral… …protective
IL23R
DLG5_e26
CARD15
OCTN 1 and 2
DLG5 R30Q
ATG16L1 T300A
NOD2/CARD15 (16q12)Caspase Recruitment Domain Family member 15
Arg702Trp
SNP8Gly908Arg
SNP12
Leu1007fsinsC
3020insC
SNP13
CARD 2
220127
NBD
273 577
LRR
1020744 1040
CARD 1
1 12428
Protein/proteininteractions
ProteinOligomerisation
Recognition PAMPs
NOD2/CARD15 NOD2/CARD15 mutationsmutations
+
αα--defensindefensinlysosimlysosim
Pattern Pattern recognition recognition receptorreceptorss((PRRsPRRs))
NOD2/CARD15 mutation prevalence
0
20
40
60
80
100
Hugot
Ahmad
Cuthbe
rtAbre
uVerm
eire
Esters
Buning
Inoue
Leon
gHeli
oArno
ttLa
katos
%
controls CD
CARD15 Mutations and relative risk for CD
29.3
-
12.9
9.8
13.6
14.4
2.4
2.8
1.7
Ahmad et al.
Esters et al.
Lakatos et al.
24.822.23.0Hampe et al.
-42.12.6Cuthbert et al.
-17.61.5Ogura et al.
44.038.03.0Hugot et al.
Compound Heterozygote
HomozygoteHeterozygote
2-6/1.000 2-8%
Ahmad et al, Gastroenterology 2002, 122: 867-874
NOD2/CARD15 and disease phenotype
PAR% of population attributable risk
Increase (proportional) in disease in individuals with mutation in comparison with those without mutation
Ileal CD Colonic CDNOD2HLAOther
Lala S et al; Gastroenterology 2003; 125: 47-57
Ogura et al; Gut 2003; 52: 1591-1597
NOD2 expression in Paneth cells
Defensins and NOD2
Wehkamp J Gut 2004; 53: 1658-64Kobayashi et al Science 2005;307:731-734
Wehkamp PNAS 2005
Fellermann K et al; Am J Hum Genet. 2006;79:439-48
0
20
40
60
80
<4 4 >4
Gene copy number
case
s (%
)
0
20
40
60
80
<4 4 >4
Gene copy number
case
s (%
)
0
20
40
60
80
<4 4 >4
Gene copy number
case
s (%
)
0
20
40
60
80
<4 4 >4
Gene copy number
case
s (%
)
Ileitis Colitis
UCIleocolitis
Copy-number variation of Human beta-defensin
1 2 3 4 5 6 7 8 9 10Chrom 4p14 4q31 4q35 9q33 1q41 4p14 Xp22 Xp22 3p21 4
Bacteriallipoprotein
MycoplasmalLipopetide
Viral dsRNA CpG DNA
motifsFlagellin
BacterialLPS
MycoplasmalLipopetide
Toll Like Receptors
Franchimont Gut 2004, Torok Clin Immunol 2004, Ouburg Gut 2005, Lakatos World J Gastro 2005, Braat J Mol Med 2005, Pierik M IBD 2006
IBD5 and OCTN (5q31-33)Novel Organic Cation Transporter
Rioux JD Nat Genet 2001; 29:223-28Peltekova Nat Genet 2004; 36: 471-5
C1672T and promoter G–207C
Risk haplotype TC
54% IBD versus 42% controls
OCTNNovel Organic Cation Transporter
• Membrane transporter for drugs and organic cations
• OCTN may also transport carnitine:– essential for metabolism of lipid– Role in transport of short chain fatty acids to mitochondria
for β-oxidation
Peltekova Nat Genet. 2004; 36: 471-5
• Rectal instillation of sodium 2-bromo-octanoate (inhibitor β-oxidation) in rats:• Weight loss and bloody diarrhea• ulcers, mucus cell depletion, vasodilation, increase of acute inflammatory cells
(Roediger et al Br J Exp 1986)
• Association of OCTN1 and RA (Tokuhiro et al Nat Genet 2003)
p=0.05
p=0.034
35
43,4 44
53,5
0
10
20
30
40
50
60
% patients
Perianal disease Fistulising disease
No TC risk haplotypeTC Homozygous
TC risk haplotype associated with fistulising disease OR 1.474 (95% CI 1.028-2.114); p=0.035
(Forward Wald log regression SPSS 12.0)
OCTN TC risk haplotype
p=0.05
p=0.034
35
43,4 44
53,5
0
10
20
30
40
50
60
% patients
Perianal disease Fistulising disease
No TC risk haplotypeTC Homozygous
OCTN TC risk haplotype?
Or non stricturing/non penetrating- colonic –no surgery??
OCTN TC risk haplotype??
DLG5 (10q23)Discs Large Homolog 5
DUF622 PDZ PDZ PDZ PDZ SH3 GK
N C
member of MAGUK (Membrane Associated Guanylate Kinase) family of scaffolding proteins
• involved in intracellular signal transduction • important in maintaining epithelial structure and integrity of barrier
G113A (R30Q)Haplotype D-tagging SNP
DLG5_e26(One of 8) Haplotype A-tagging SNPs
Stoll M et al Nat Genet. 2004; 36: 476-80
Stoll M et al Nat Genet. 2004; 36: 476-80
DLG5 R30Q
25
17
0
10
20
30
%
CD (n=538) Controls (n=548)
p=0.001
OR=1.6
Daly M et al Eur J Hum Genet 2005;
0.5 1 1.5 2 2.5
Canada/Italy
UK
Canada/Italy
Estimated OR for R30Q
OR=1.25
Not confirmed in other studies:
Lakatos PL IBD 2006, Noble CL Gut 2005, Torok HP Gut 2005, Vermeire S Gastroenterol 2005, Waller S DDW 2005
From linkage analyses usingmicrosatellites…
…to SNP association genome scans
NOD1 (CARD4) (7p14)IL23R (1p31)- the 300.000 SNP story
Duerr RH Science 2006
IL23R (1p31)
Arg381Gln
1,9
7
012345678
CD controls
All
freq
%
Confirmed in other studies:
Cummings Inflamm Bowel Dis 2007 + rs7517847 (OR 0.65, 95%CI: 0.56-0.75)
Tremelling Gastroenterology 2007
ATG16L1 (2q)Autophagy-related 16-like 1 gene
OR G allele: 1.35-1.45, GG gen: 1.71-1.77
Confirmed in other studies:
Rioux Nat Genet 2007, Prescott NJ Gastroenterology 2007
Cummings Inflamm Bowel Dis 2007 no association to NOD2/IL23,IBD5 & phenotype
Genetic markers studied: TUCAN (CARD8), NOD1, NOD2, IBD5, TNFSF15
Gene chip for the diagnosis of IBD? Genetics in combination with environmental factors
McGowern Gastroenterology 2006
Predicting respone to medical therapy: A role for pharmacogenetics?
• TNFa, TNFR, NOD2, or NAT1-2not associated to response to steroids, azathioprine, mesalasine/sulphasalazineor infliximab
• MDR1:response to AZA and steroid dependency?
• IL10 1082AA: steroid dependency?
• DLG5 R30Q: resistance to steroids?
• FAS 843C/T and caspase9 93C/T: response to IFX? (concomittant AZA)
• TPMT: azathioprine!
So…genotype is also complex
Genetics start to unravel the complex phenotype of IBD
IBD9IBD5
IBD4
IBD3
IBD2
IBD1
IBD7
IBD6
DLG5
CARD15
OCTN
NOD1
IL23R
Genetics in clinical practice
NOD2/CARD15 is until now the most important NOD2/CARD15 is until now the most important genetic factor implicated in the disease susceptibility togenetic factor implicated in the disease susceptibility toCDCDConfConfiirmsrms the hypothesis that the hypothesis that dysregulateddysregulated response of response of the mucosal immune system toward microbial antigens the mucosal immune system toward microbial antigens may lead to the development of may lead to the development of CDCDClinical utilization:Clinical utilization:
Genetic testing of NOD2 may be an additional confirmatory Genetic testing of NOD2 may be an additional confirmatory factor in making the diagnosis and predicting disease factor in making the diagnosis and predicting disease phenotypephenotype to some extent (e.g. early onset , ileal involvement, to some extent (e.g. early onset , ileal involvement, stenostenossinging disease, increased need for surgery?)disease, increased need for surgery?)
‘‘‘‘The conclusions with respect to The conclusions with respect to the the NOD2/CARD15NOD2/CARD15
experience in inflammatory bowel experience in inflammatory bowel disease have led todisease have led to
recommendations with respect to recommendations with respect to clinical classificationclinical classification ofof
DiseaseDisease””
Biomarkers: for the diagnosis
Serologic markers in IBD
++ (15-30%)PancreasPAB
-+ (20-40%)glycan (Chitobioside)ACCA
-+ (20-50%)flagellincBir1
glycan (Laminaribioside)
Pseudomonas fluorescens
Outer Membrane porin
Neutrophils
Mannose of Saccharomycescerevisiae
Directed against
+ (40-60%)+ (UC-like CD)
pANCA
-+ (40-60%)ASCA
-+ (20-40%)I2
++ (20-40%)Omp (C)
-+ (20-40%)ALCA
UCCDAntibody
Mow S et al, Gastroenterology 2004; Targan SR et al, Gastroenterology 2005
A panel of markers may improve sensitivityand approach diagnostic utility
Dotan I et al, Gastroenterology2006; 131: 366-78
↔ Reduced specificity
↔ Considerably expensive
↔ Not widely available
S Joossens et al, Gastroenterology, 2002, 122: 1242-1247
n CD UC IC
ASCA+/ pANCA- 26 (26.8) 8 (30.8) 2 (7.7) 16 (61.5)
ASCA-/pANCA+ 20 (20.6) 4 (20) 7 (35) 9 (45)
ASCA+/ pANCA+ 4 (4.1) 2 (50) 1 (25) 1 (25)
ASCA-/pANCA- 47 (48.5) 3 (6.4) 4 (8.4) 40 (85.1)
Total 97 (100) 17 (17.5) 14 (14.4) 66 (68.1)
P<0.001
(Leuven n=30, Lille n=37, Vienna n=30)
Serologic markers in patients withcolitis type unclassified (IBDU)
Dubinsky M et al Am J Gastroenterol 2001
ASCA and ANCA ELISA
ASCA ELISA - pANCA 3-steps
Serologic markers may prioritize diagnostic work up in children
Dubinsky M et al Am J Gastroenterol 2001
↓
4/128 would have receivedunnecessary investigations
↑
Delayeddiagnosis
Novel algorythmic analysis of serology data for predicting IBD
Lichtenstein Gastroenterology 2007;132: A175 (S1106)
•Sophisticated computer-aided system developed from a sequential analysis of serologic assay results using two stage statistical classifiers including a neural network
•Prevalence of IBD in validation cohort (n=500): 59%•Overall accuracy of algorithm: 92%
939890NPV
988996PPV
989795Specificity
889393Sensitivity
CD (n=188)UC (n=105)IBD
Israeli Defense Corp serumrepository
32 persons with CD from whomserum before the diagnosis
ASCA in 10/32 (32%) before diagnosis
mean interval between ASCA detection and diagnosis 38 M
Israeli E, et al. Gut 2005
‘Subclinical’ disease: is serology preceeding disease?ASCA before onset of Crohn’s disease
Time before/after CD diagnosis (months)
0
70
80
90
100
-60 -48 -36 -24 -12 6 18
No
with
ASC
A p
ositi
ve (%
)
40
50
60
10
20
30
-54 -42 -30 -18 -6 0 12
Diagnosis
Is there a genetic susceptibility for the seroreactivity in IBD?
Antimicrobial peptidesASCA, anti-OmpC, Anti-
I2, cBir1 flagellin
NOD2/CARD15, NOD1, TLR4, ATG16L1, …
ASCA, I2, CBir1, OmpC
Devlin S et al, Gastroenterology 2007
9,7
10,4
11,3
9
9,5
10
10,5
11
11,5
12
No Variant 1 Variant 2 Variants
NOD2 Variant Status
Mea
n Q
uart
ile S
um
732 CD
p=0.002
Antimicrobial responses as consequence of geneticdefects in innate immune system?
9,7
10,8
9
9,5
10
10,5
11
11,5
12
No Variant Any Variant
NOD2 Variant Status
Mea
n Q
uart
ile S
um 220 Healthy relatives
p=0.02
9,7
10,7
9
9,5
10
10,5
11
11,5
12
No Variant Any Variant
NOD2 Variant Status
Mea
n Q
uart
ile S
um 200 Healthy controls
p=0.07
Antimicrobial responses due to genetic defects in innate immune system?
Henckaerts L al; submitted
51%
0%
25%
50%
75%
100%
CARD15 0
64%
CARD15 1
72%
CARD15 2
n = 238 n = 187 n = 68
p < 0.0001
0%
25%
50%
75%
100%
34%
TLR 4 0
25%
TLR4 1
9%
TLR4 2
n=230 n=28 n=1
p = 0.03
gASCA
ACCA
0102030405060708090
100
0 1 2
Papp M Inflamm Bowel Dis 2007
% p<0.0001
ASCA
CARD15
• n=82 adults with abdominal symptoms – St Mark’s hospital• All received clinical examination, rectal biopsy, ESR, CRP,
α1 glycoprotein
Shine et al; Clin Chim Acta 1985; 148: 105-9
0%50%100%increased CRP
Functional boweldisorder (n=41)
UC (n=22)
CD (n=19)
Serum markers of inflammation in IBD
Serum markers of inflammation in IBD• n=91 children - mean 11yr (3M-18yr)• Referred for endoscopy with symptoms of abdominal pain, diarrhea,
rectal bleeding, weight loss or mouth ulceration for minimum of 3 M• All underwent endoscopy, blood tests and SBFT
Beattie et al; Arch Dis Childhood 1995; 73: 354-5
*TBC (2), IC (3), lymphoid nodular hyperplasia (3)
284214424228
6006091
120012088
30231570608
168535881000
Hb <10 g/dLESR >25 mm/hourAlbumin <36 g/dLPlatelet >400 x 109/lCRP >5 mg/LAll normal
Others* (n=7)
Normal (n=37)
Polyps(n=8)
Ulcerativecolitis(n=13)
Crohn’s disease(n=26)
Investigation
Tibble et al, Gut 2000
Stool markers: calprotectin• Non-invasive 36 kDa calcium and zinc
binding protein (S100A8/A9)
• Neutrophil marker: 60% of cytosolicproteins in granulocytes
• Presence in faeces proportional to neutrophil infiltration in mucosa and shedding in the gut lumen
• Antibacterial, antifungal activity
• Resistant to degradation: stable for >1 week at room temp
• Non-specific (neoplasia, infections, polyps)
Serum markers of inflammation in IBD
Tibble Gut 2000
Serum markers of inflammation in IBD
Von Roon AC AJG 2007
-5983 patients-For IBD: cut off 100 ug/g instead of 50 ug/g
-Sensitivity 95%-Specificity 91%-ROC AUC 95%
Faecal S100A12(Calgranulin C, EN-RAGE)
De Jong NS et al; Inflamm Bowel Dis 2006;12:566-72Sidler MA, Gastroenterology DDW 2007897981CRP (3mg/L)
919374ESR (15mm/h)
9567100Calprotectin(50 mg/kg)
999797S100A12 (10mg/kg)
ROC-AUC (%)
Specificity (%)Sensitivity (%)Children- member of the S100 family of calcium-binding proteins in granulocytes- exhibits proinflammatory functions, including potent chemotactic activity- ligand for the receptor for advanced glycation end products (RAGE)
Faecal lactoferrin
Walker TR et al; J Pediatr Gastroenterol Nutr. 2007; 44:414-22
• 76 kDa iron binding glycoprotein • Major component of secondary granules of PMN neutrophils• Released during inflammatory process – role in innate immunity
IBD vs healthy controls
Fagerberg UL et al J Pediatri Gastroenterol Nutr 2003; 37: 468-72
Fagerberg UL et al J Pediatr Gastroenterol Nutr. 2005; 40:450-5
Stool markers in pediatric IBD• N=36 children with symptoms and suspected inflammation of the colon
• Correlation between calprotectin in stools and endoscopic findings
n=14 n=22IBD in 20/22 children
Sens 95%, spec 93%
PPV 95%, NPV 93%
Are biomarkers helpful in predicting disease course andcomplications?
FibrostenosisSB surgery Perforating
SB diseasefibrostenosisperforatingSB surgeryNo NOD2
Mow S et al, Gastroenterology 2004; Targan SR et al, Gastroenterology 2005, Papp, Inflamm Bowel Dis 2007
Anti-flagellin
Fibrostenosisperforating
NOD2
Serological markers predict complicated disease?
83,2
71,7
56,0
42,0
0%
25%
50%
75%
100%
Group A (n=181)
Group B (n=200)
Group C (n=184)
Group D(n=173)
% w
ith
com
plic
ated
dis
ease
Score
< 1.5
Score
1.5 or 2.0
Score
2.5 or 3.0
Score
> 3.0
OR: 1.76p: .006
OR: 2.00p: .001
OR: 1.96p: .010
Serological markers predict complicated disease?
Ferrante M et al; Gut apr 24 2007 epubPapp M Inflamm Bowel Dis 2007Dubinsky Am J GE 2006
Adults: gASCA, AMCA, ALCA, ACCA, Omp
65 48 79 57 78 800
102030405060708090
100
0 1 2
Ilealnon-inflamm
%
Serology marker positivity
ASCA, Omp
OR: 1.92-4.28For 0 vs 2
Children: ASCA, Omp, I2, CBir1
Serological markers predict aggressive disease?
Dubinsky M et al; Am J Gastroenterol 2006; 101:360-367
ASCA, OmpC, cBir1, I2
SatsangiSatsangi J J GGutut 2006;55:7492006;55:749--53.53.
‘‘the assessment of the role of serological markers at the present time, the Working Party determined that the use of these markers for diagnosis is not currently justified, given the limited sensitivity of available markers’’
The clinical utility of serological markers?
Assessment of disease activity?
Clinical disease and the CRP response
SLE, Scleroderma, Dermatomyositis, Ulcerative colitis, Leukemia, Graft-vs.-host disease
Modest or absent CRP response
Bacterial, Fungal, Mycobacterial, Viral (severe or systemic)Rheumatic fever, Erythema nodosum
Rheumatoid arthritis, Ankylosingspondylitis, psoriatic arthritis, systemic vasculitis, PMR, Crohn’s diseaseObesity, atherosclerosis
Myocardial infection, Tumourembolization, Acute pancreatitisSurgery, burns, fracturesLymphoma, carcinoma, sarcoma
Major CRP responseInfections
Hypersensitivity complication of infection
Inflammatory disease
Metabolic diseases
Tissue necrosis
TraumaNeoplasia
CRP in IBD
0102030405060708090
Mild Moderate Severe
Median CRP mg/L
CD (n=64)UC (n=50)
Fagan EA et al. Eur J Clin Invest 1982; 12: 351-359
prospective
Correlation of CRP with clinical, radiographic and endoscopic activity in IBD
UC (n=43)
OR (95% CI)Logistic regressionIncreased CRP associated with:
NSNS
p=0.029NS
4.5 (1.1-18)4.1 (1.6-11)10 (1.0-97)
NS
Clinical activityEndoscopic activitySevere inflammation on BiopsyAbnormal Radiographic findings
CD (n=104)
Solem CA et al; Inflamm Bowel Dis 2005; 11:8:707-712
25%54%CRP elevated
75%46%CRP normal
Ileocolonoscopy does not show active CD (n=32)
Ileocolonoscopy shows active CD (n=72)
CRP or ESR ?
Fagan EA et al. Eur J Clin Invest 1982; 12: 351-359
Mild Moderate0
20
40
60
80
Severe
100
120
140
160 CRP
SevereMild Moderate
0
20
40
60
80
100 ESR
CRP or ESR ?
Consigny et al Inflamm Bowel Dis 2006;12:551
10.5 (2.3-48.1)CRP >20 mg/L only
9.9 (3.3-29.7)Both6.1 (1.9-18.9)ESR >15 mm only
RR of relapse (95%CI)
•71 CD patients in medical remission
•CRP>20mg/L and ESR >15 mm/h were selected as markers predictiveof relapse
•A binary biological predictive score was derived: "negative" when both were lower thantheir limits, "positive" when otherwise
•Sensitivity 89% •specificity was 43%
Faecal calprotectin
Roseth A et al Digestion 1997; 58: 176-80Roseth A et al Scand J Gastroenterol 2004; 39: 1017-1020
Predicting relapse?
Faecal calprotectin and risk for relapse
Tibble Gastroenterology 2000
0
RR 10.6 (CD)
RR 13.4 (UC)
• 43 CD
• 37 UC
• In remission for 1-4 months
25 (58%) relapse over period of 12 months
19 (51%) relapsed over period of 12 months
Proportion of patients without
a relapse
Time (Months)
UC Calprotectin <50mg/L
CD Calprotectin <50 mg/L
UC Calprotectin >50 mg/L
CD Calprotectin >50 mg/L
3 6 9 120
0.5
1
0.25
0.75
Faecal calprotectin and risk for relapse
Costa F Gut 2005
• 38 CD
• 41 UC
15 (39%) relapse over period of 12 months
19 (46%) relapsed over period of 12 months
Risk of relapse after withdrawal of azathioprine
Lemann M et al Gastroenterology 2005; 128: 1812-1818
84 CD in remission>42 months with AZA
AZA n=40
Placebo n=43
% of clinical relapse at 18 months?
21.3%
7.9%
Calprotectin predicts endoscopic post-surgical recurrence in asymptomatic CD
Orlando Eur Rev Med Pharmacol 2006;10:17
•39/50 patients evaluated at 1 yrpost-op by colonoscopy
•19 pts had endoscopic recurrence
•Calprotectin>200mg/L improvedsensitivity when compared to US
•Detection of calprotectin>200mg/L at third month aftersurgery represents an indicationto endoscopic examination
7563>200
8547>250
5578>150
4578>100
2594>50
Specificity (%)Sensitivity (%)CalprotectinCut-off
Conclusions
• Genetic markers:– provide evidence that altered NOD2/CARD15 (or TLR4)-mediated bacterial
sensing of normal commensal flora and mucosal permeability changes may be key mechanisms in the pathogenesis of IBD
– low sensitivity limited diagnostic value (targeted chip?)– Associated with disease phenotype (e.g. CARD15 and ileal/stenosing,
OCTN and perianal, HD and colon)
• Serological markers:– loss of tolerance to bacterial antigens in IBD – Low sensitivity: little value in diagnosis (panel-algorithm of markers?) – Associated with complicated (ileal) disease behaviour
• Other laboratory and stool markers:– Useful tools and should be part of global management of IBD– CRP: Good correlation for CD with disease activity-predicting response-
relapse, correlation less for UC except for severe extensive colitis – Stoolmarkers: more specific to detect gut inflammation in the colon