1. Prognostication in IBD: Can We Really Predict the Future or Are We Just Describing the Past?
Raymond Cross, MD, MS, AGAF
Associate Professor of Medicine
Director of the Inflammatory Bowel Disease Program
University of Maryland School of Medicine
7/19/14

2. Current Need for Risk Stratification in IBD
IBD, especially CD, are heterogeneous diseases
Not appropriate to treat every patient the same
Despite improved outcomes with early biologic therapy with IS, this approach may not be appropriate for every patient
Overtreatment of patients with mild disease
Increased rates of adverse events (albeit small)
Cost
There is often a poor correlation between symptoms and disease activity
Personalized approaches which include risk stratification are needed to individualize therapy

3. What clinical variables are associated with adverse outcomes in IBD?

4. Clinical Predictors of Complicated Clinical Course in CD
5-year clinical course after diagnosis
Variable
Nondisabling, % (n=166)
Disabling, % (n=957)
P-value*
Age at onset
<40 years*
77.1
87.7
.0004
≥40 years
22.9
12.3
Location of disease
Small bowel only
44.6
32.8
.002
Small bowel + colon
25.9
39.4
Colon only
29.5
27.8
Smoking status
Smoker
50.3
57.4
.09
Ex- or nonsmoker
49.7
42.6
Perianal lesions at diagnosis
Yes*
17.5
26.4
.01 No
82.5
73.6
Required steroids for tx first flare
37.3
65.2
.0001
62.7
34.8
Beaugerie L, et al. Gastroenterology. 2006 4

5. Predictors of Complicated CD
Complicated CD defined as either stricturing or penetrating disease
Patients with complicated CD were
Younger at diagnosis (24 years vs. 33 years, p<.001)
Current smokers (36% vs. 15%, p<.01)
Ileal disease (39% vs. 19%, p<.01)
Exposed to steroids in 1st year after diagnosis (77% vs. 56%, p<.01)
Less likely to be exposed to biologics (6% vs. 20%, p=.01)
Patil, S., et al. (2012). Gastroenterology 142(5): S667-S667

6. Variables Significantly Associated with Early Surgery in Crohn’s Disease
Factor
Odds Ratio
95% CI
Oral corticosteroid use in 1st
6 months vs. none
3.79
( )
Current smoker
vs. nonsmoker
3.09
( )
Ileal localization only
vs. all others
2.22
( )
Nausea/vomiting vs. none
2.07
( )
Abdominal pain vs. none
1.82
( )
Colonic localization only
0.27
( )
Sands B., et al. Am J Gastroenterol 6

7. Clinical Predictors of Severe and “Very Severe” Crohn’s Disease
Severe CD (58% of subjects)
Age <40 years at onset
Presence of perianal lesions at diagnosis
Very severe CD
(37% of subjects)
Stricturing or intra-abdominal fistulizing at diagnosis
Fever at diagnosis
Weight loss >5 kg at diagnosis
Increased platelet count at diagnosis
-Criteria for severe CD same as those defined by Beaugerie
-Very severe CD defined as: >70 cm intestinal resection; >2 intestinal resections; colectomy; stoma; complex perianal disease
*P<0.001; †P<0.01; ‡P<0.05
Loly, C., et al. (2008). Scand J Gastroenterol 7

8. Variables Associated with Severe CD
Early age at diagnosis
Early treatment with corticosteroids
Perianal involvement
Smoking
Disease location
Ileal vs. colonic
Upper GI tract involvement?
Complicated disease at diagnosis

9. Are biologic markers associated with a more severe disease course in IBD?

10. High Fecal Calprotectin is Associated with Risk of Relapse in IBD
43 CD
37 UC
In remission for 1-4 months
25 (58%) relapsed over period of 12 months
19 (51%) relapsed over period of 12 months
Proportion of patients without a relapse 1
UC calprotectin <50mg/L
UC calprotectin >50 mg/L
CD calprotectin <50 mg/L
CD calprotectin >50 mg/L
RR 10.6 (CD)
RR 13.4 (UC) 3 6 9 12
RR, relative risk
Tibble, JA., et al. Gastroenterology. 2000
Time (months) 10

11. FCP as Marker of Disease Activity in Patients Taking Infliximab for UC
UC patients in remission on IFX 5 mg/kg q8w (n=113)
FCP measured monthly for 1 year
Deep remission = normal endoscopy at baseline and week 52 + Mayo score < 3
27% were in deep remission
FCP levels highly correlate with disease activity
FCP <50 mg/kg at all time points in patients in deep remission
Median FCP 477 mg/kg in patients with a flare
Two consecutive levels >300 mg/kg predicted a flare
De Vos M et al. ECCO Abstract no OP 07.

12. FCP and Fecal Lactoferrin Predict Endoscopic Disease Activity
FCP and FL correlate significantly with the Crohn’s Disease Index of Severity (CDEIS)
Measure
Sensitivity
Specificity
PPV
NPV
CDAI>150
27%
94%
91%
40%
CRP>5mg/l
48%
FCP >200mcg/g
70%
92%
61%
FL >10mcg/g
66%
59%
Sipponen T, et al. Inflamm Bowel Dis 2009

13. Antibody Responses to Microbial Antigens Predict Clinical Outcomes
Variable
Small bowel disease
Fibrostenosis
Internal perforating
Small bowel surgery
UC-like
Anti-I2 NS
0.027
0.01
Anti-OmpC
<0.006
ASCA
0.023
<0.001
<0.001¥
pANCA
0.013¥
<0.002¥
NOD2
0.003
<0.008¥
¥negative association
Mow, W. S, et al. (2004). Gastroenterology

14. Risk of Disease Progression Increases With Increased Immune Responses
CLINICAL PHENOTYPE
NUMBER OF ANTIBODIES TOWARD MICROBIAL ANTIGENS*
P trend
ODDS RATIO (3 vs 0)
95% CI (3 vs 0) 1 2 3
Fibrostenosing (%)
23.0
50.0
66.7
72.0
<0.001
8.6
4.0–18.9
Internal perforating (%)
27.9
27.5
42.5
58.7
3.7
1.8–7.6
Small bowel surgery (%)
57.5
*Number of antibodies is the cumulative total of microbial antigen responses to I2, OmpC (outer membrane porin C), and ASCA (anti-Saccharomyces cerevisiae). Results are irrespective of pANCA and NOD/CARD15 status.
Mow, W. S, et al. (2004). Gastroenterology 14

15. Increased Immune Reactivity is Associated with Complications in CD
Mow WS, et al. Gastroenterology. 2004

16. Immunoreactivity Predicts Rapid Progression to Complications and Surgery
Dubinsky MC, et al. Clin Gastroenterol Hepatol. 2008

17. Requirement for Surgery in CD According to NOD2/CARD15 Gene Status
Results of Kaplan-Meier Analysis: Probability of Remaining Free of Surgery According to NOD2/CARD15 Gene Status
Presence of NOD2/CARD15 variants was associated with:
More frequent surgery for stricture
Earlier requirement of surgery
More frequent surgical recurrence
Earlier requirement for reoperation
Months 20 80
100
120 40 60
0.0
1.00
0.6
0.2
Cumulative Survival
0.4
0.8
No variant
NOD2/CARD15 gene variants
Alvarez-Lobos M, et al. Ann Surg. 2005

18. Incidence of Pouchitis According to Level of pANCA
Fleshner PR, et al. Gut. 2001
K25RGU121005

19. Time after IPAA (months)
Effect of Anti-CBir1 Expression and pANCA Level on Cumulative Incidence of Chronic Pouchitis
Time after IPAA (months)
Fleshner PR, et al. Clin Gastroenterol Hepatol. 2008

20. Biologic Markers Can Predict Relapse and Complications in IBD
FCP correlates with endoscopic disease activity
Persistently elevated levels predict relapses
Qualitative and quantitative responses to microbial antigens are associated with complications in CD and UC
Mutations in NOD2 are associated with higher rates of surgery

21. Can therapeutic drug levels predict future outcomes?

22. Detectable IFX Trough Levels are Associated with Improved Outcomes
Maser, E. A., et al. (2006). Clin Gastroenterol Hepatol

23. Detectable IFX Trough Levels are Associated with Improved Outcomes
CRP 2.0 in patients with detectable trough vs in
those with negative levels
Maser, E. A., et al. (2006). Clin Gastroenterol Hepatol

24. Use of Therapeutic Drug Levels to Guide Changes in Therapy
Response to Test
Clinical Response
P value
Detectable HACA
Increase IFX
17%
P<0.004
Change Anti-TNF
92%
Subtherapeutic concentration
86%
P<0.016
33%
Afif W., et al. Am J Gastroenterol 2010

25. Trough and ATI Predict Persistence with IFX in IBD
Vande Casteele N. Am J Gastroenterol 2013

26. Trough and ATI Predict Persistence with IFX in IBD
ATI were present in 59% of patients; 72% were persistent and 28% were transient
Vande Casteele N. Am J Gastroenterol 2013

27. Trough and ATI Predict Persistence with IFX in IBD
IFX trough level < 2.2 ug/ml at week 14 predicted IFX LOR and hypersensitivity reaction
ATI were present in 59% of patients; 72% were persistent and 28% were transient
Vande Casteele N. Am J Gastroenterol 2013

28. Trough Levels of IFX Predict Outcomes in UC
Seow, C. H., et al. (2009). Gut.

29. Can endoscopic findings predict the subsequent disease course?

30. Definition of Mucosal Healing
Definitions of MH not uniform
Interobserver variability
Reasonable definition would be total disappearance of all mucosal ulcers
Erythema and distorted mucosal vascular pattern not included
Is improvement in endoscopic activity just as important?

31. Risk of Colectomy According to the Presence of Severe Endoscopic Lesions
Probability of Colectomy (%)
*Extensive and deep ulcerations on index colonoscopy.
Allez M, et al. Am J Gastroenterol. 2002
K25RGU121005 31

32. Long-term Outcomes Following Mucosal Healing: Norway
No mucosal healing
Froslie KF, et al. Gastroenterology 2007

33. MH After Therapy Predicts Improved Outcomes
Baert, F., et al. (2009). Gastroenterology.

34. Is Complete MH Needed in CD?
Schnitzler, F., et al. Inflamm Bowel Dis 2009

35. Curatio PowerPoint Template
12/27/ :11 AM
EXTEND: Patients Who Achieved Deep Remission* with ADA at Week 12 and Hospitalization Rates
All-cause hospitalization through Week 52
CD-related hospitalization through Week 52 20 20 17 15 15
All hospitalization (%) 10
CD-related hospitalization (%) 10 9 5 5
0/11
9/53
0/11
5/53
Deep remission* (Week 12)
Non-deep remission* (Week 12)
Deep remission*
(Week 12)
Non-deep remission*
(Week 12)
* Deep remission defined as clinical remission (CDAI <150) and complete mucosal healing in EXTEND
Colombel JF et al. Gut. 2010;59(Suppl 3):A80.

36. Long Term Outcomes Following MH in Patients with UC
Adapted from Froslie KF, et al. Gastroenterology 2007

37. MH after Steroids Associated with Improved Clinical Outcomes in UC
Variable
CR & ER
CR only
No response
General relapse
83%
92%
100%
Systemic relapse
55%
72%
91%
Hospitalization
25%
49%
64%
Immune suppressant 5%
26%
54%
Colectomy 3%
18%
17%
*Mean follow up 51months
§76% followed for 5 years
Ardizzone, S.,et al. (2011). Clin Gastroenterol Hepatol

38. Week 8 Endoscopy Correlates With Steroid-free Remission Rates at Week 30 in ACT 1 and ACT 2
Week 8 Mayo endoscopy score
Median steroid dose
Steroid-free remission rates
P value
IFX Arm 46
<0.0001 1
4.5 34 2 10 11 3
6.5
Colombel, JF, et al. Gastroenterology 2011

39. MH and Time to Colectomy in IFX-treated Patients
0 = NORMAL
1 = MILD
2 = MODERATE
3 = SEVERE
Colombel JF et al. Gastroenterology. 2011

40. MH is Associated with Improved Outcomes
Achieving MH (or improvement in endoscopic activity) is associated with decreased:
Flares
Surgery
Hospitalizations
Should MH be the target of treatment?
Cannot achieve MH in a significant proportion
Partial MH may be good enough
Limited prospective studies evaluating outcomes in those treated with MH as treatment target

41. Conclusion
Currently available clinical, biologic, genetic and endoscopic markers are helpful to identify patients at risk for complications of IBD
In the future, improved diagnostics will help identify at-risk (at-higher risk) IBD patients that will benefit from aggressive medical intervention
Anticipate that diagnostics will also help to identify the best (or worse) treatment for a patient