““Top‐down”
Top‐down” therapy:
Is the evidence strong enough?
Eugeni Domènech, MD, PhD
IBD Unit
Gastroenterology & Hepatology Department
Hospital Universitari Germans Trias i Pujol
Badalona, Catalonia. Spain.
Falk Symposium 168
IBD in different age groups
Madrid, March 27‐28, 2009
Short‐term outcome in Crohn’s disease
patients requiring systemic steroids
Disease activity
1‐year outcome
Steroid dependency 22–23%
Surgery
35%
1 mo
1 year
Faubion WA, et al. Gastroenterology 2001
Ho G‐T, et al. Aliment Pharmacol Ther 2006
Long‐term outcome in Crohn’s disease.
Stenosing and penetrating complications
100
100
% patients
cumulative probability (%)
B3 penetrating
80
60
B2 stenosing
40
20
0
B1 inflammatory
0
1
3
5
10
15
time from diagnosis (years)
Louis E, et al. Gut 2001
20
25
penetrating
80
60
40
inflammatory
20
stenosing
0
0
24
48
72
96 120 144 168 192 216 240
time from diagnosis (months)
Cosnes J, et al. Inflamm Bowel Dis 2002
Long‐term outcome in Crohn’s disease.
Requirements of intestinal resection
100
Cumulative probability of intestinal resection
80
60
40
20
1
5
10
Years from diagnosis
Hospital‐based incident cases 1980‐1999 (n=480). Tavarela‐Veloso F, et al. Inflamm Bowel Dis 2001
Population‐based incident cases 1991‐2003 (n=476). Romberg‐Camps MJL, et al. Am J Gastroenterol 2009
Long‐term outcome in Crohn’s disease.
100
Cumulative probability of “all type” recurrences
80
60
40
20
1
5
10
Years from diagnosis
Population‐based incident cases 1991‐2003 (n=476).
Romberg‐Camps MJL, et al. Am J Gastroenterol 2009
Short‐term outcome in patients with
mild‐to‐moderate inflammatory activity
RCT comparing budesonide to placebo for maintenance of remission
in CD patients with prior mild‐to moderate activity
Hanauer SB, et al. Aliment Pharmacol Ther 2005
Natural history of Crohn’s disease: Early control of
disease and recurring flares
Observational study of 480 consecutive CD patients followed for 20 years
(mean: 7.2 years) from disease diagnosis (1980–1999)
Tavarela‐Veloso FT, et al. Inflamm Bowel Dis 2001
Drug efficacy depends on disease duration
Prospective follow‐up after a single IFX infusion of 15 pediatric patients with refractory CD.
Patients without relapse
100
early CD (<2 years)
late CD (>2 years)
75
50
25
12
24
36
48
Weeks following IFX infusion
Kugathasan S, et al. Am J Gastroenterol 2000
Drug efficacy depends on disease duration
AZATHIOPRINE as maintenance
of steroid‐induced remission
Median
disease
duration
91%
42%
32%
<8 weeks
2.6 years
4 years
1
Markowitz J, et al. Gastroenterology 2000
2 Candy S, et al. Gut 1995
3
Lémann M, et al. Gastroenterology 2006
INFLIXIMAB for induction of
remission (3 infusions)
64%
2 weeks
1
59%
1.6 years
42%
8 years
D’Haens G, et al. Lancet 2008 (SUTD)
2 Hyams J, et al. Gastroenterology 2007 (REACH)
3 Rutgeerts P, et al. Gastroenterology 2004 (ACCENT‐I)
Immune responses vary with time in CD
Cytokine production by T‐cells from colonic mucosal biopsies in children
with early CD (at diagnosis) or late CD ( at least 5 years from diagnosis).
Kugathasan S, et al. Gut 2007
Treatment strategies and disease prognosis
Good prognosis
Limited efficacy
Treatment goal:
symptom abatement
Minimal risk of AE
Poor prognosis
Treatment goals:
Prevent disease complications
Avoid admissions, surgeries, disability
Great efficacy
High risk of AE
Top‐down strategies: Potential Pros and Cons
Pros
Cons
Longer remission periods
Less relapses
Less hospital admissions
Lower rates of intestinal resections
Overtreatment
Safety concerns: infections &
malignancies
Economic concerns
CHANGING NATURAL HISTORY?
DO THEY REALLY CHANGE
NATURAL HISTORY?
Early introduction of thiopurines
RCT comparing AZA vs PBO in new onset pediatric CD (<8 weeks). Initial course with
40mg PDN (n=55). 18‐months follow‐up.
Remission without steroids
7500
*p<0.03
5000
*
*
*
*
*
*
2500
75
50
placebo
25
0
0
placebo
6-MP
100
Remission(%)
Cumulative PDN dose (mg)
Steroid exposition
3
6-MP
6
12
months
18
0
100
200
300
400
500
600
days
Markowitz J, et al. Gastroenterology 2000
Early introduction of thiopurines
Retrospective study. Pediatric new‐onset CD cohort (The Netherlands).
Clinical outcome after first remission with corticosteroids depending on the
early introduction of AZA or not (n=58).
Jaspers GJ, et al. Inflamm Bowel Dis 2006
Impact of thiopurines on CD natural history
Retrospective study of a poblational incident cohort of 404 pediatric CD patients
(1988‐2002). Assessment of therapeutic requirements and development of disease
complications.
Risk factors for surgery in pediatric CD patients
(multivariable Cox model)
B2 behaviour
Corticosteroids
AZATHIOPRINE
Hazards ratio
95% CI
P value
2.54
2.98
0.51
1.58‐4.01
1.64‐5.41
0.33‐0.78
<0.01
<0.01
<0.01
Vernier‐Massouille G, et al. Gastroenterology 2008
Conventional therapy vs ‘top‐down’ in new onset CD
RCT comparing conventional therapy or “STEP‐UP” (PDN→PDN → AZA →IFX) vs intensive therapy or
“TOP‐DOWN” (IFX+AZA →IFX →PDN) (n=133). All patients were naïve for immunomodulators and
biologicals. 2‐years follow‐up.
D’Haens G, et al. Lancet 2008
% patients given antimetabolites
Is it necessary to early introduce AZA in all CD patients ?
“Top‐down”
100
overtreated
80
60
“Step‐up”
undertreated
40
20
26
52
78
104
D’Haens G, et al. Lancet 2008
… but still a great proportion relapse!!
100
Probability to remain in remission
80
60
“Top‐down”
40
P=0.031
“Step‐up”
20
26
52
78
104
D’Haens G, et al. Lancet 2008
Mucosal healing
SUTD endoscopic substudy after 2 years.
Proportion of patients with mucosal healing
73%
30%
Top –down
(IFX+AZA)
n=26
Step‐up
(PDN+/‐AZA)
n=23
D’Haens G, et al. Lancet 2008
Prognostic value of mucosal healing in new onset CD
Follow‐up (4 years) of 42 CD patients included in the SUTD study.
Patients remaining in clinical
remission after 4 years
P = 0,004
15/22
7/19
Baert F, et al. DDW 2008
What does it mean “top‐down” therapy?
“Top‐down” therapy:
early use of immunomodulators and biological agents.
Shergill & Terdiman. World J Gastroenterol 2008
“Top‐down” therapy:
use of highly effective, but potentially more toxic, treatment
strategies early in the course of a chronic disease, in order to
prevent disease progression and complications.
Armuzzi A, et al.Dig Liv Dis 2008
MAINTENANCE
INDUCTION
How many ‘top‐down’ strategies are there?
‘top‐down’, ‘middle‐up’, or ‘accelerated step‐up’?
surgery
biologicals
thiopurines / MTX
prednisone / budesonide
5‐ASA
antibiotics
Conclusions
— Conventional treatment strategies in CD are
associated to a high rate of ‘complicated disease’
within the first 5‐10 years from diagnosis.
— The earlier immunomodulators and/or biologicals
are used, the better they work.
— To demonstrate that any drug changes CD’s natural
history, long‐term follow‐up studies are needed, but
maybe these could be replaced by mucosal healing
assessment.
Conclusions
— By now, only the early introduction of thiopurines
has demonstrated to change natural history by
reducing the risk of disease relapses and surgical
requirements. Because useful predictors of
disease outcomes are still lacking, early
introduction of thiopurines may overtreat 10 to
30% of patients.
— There are many potential ‘top‐down’ strategies,
but most of them have not been evaluated. Their
evaluation and comparison in RCTs are warranted.
Lessons from post‐operative recurrence:
mucosal lesions and risk of clinical recurrence
I0 + I1
I2
I3
I4
Rutgeerts P, et al. Gastroenterology 1990
Long‐term results of ileocecal resection for CD
139 CD patients undergoing (first) ileocecal resection for CD between 1980‐2000. None
received biologicals prior to resection. Prospectively maintained database (Ireland).
‘Disease recurrence’ = symptoms + endoscopic or radiologic evidence of recurrence.
Cullen G, et al. Inflamm Bowel Dis 2007
How many ‘top‐down’ strategies are there?
Ongoing RCT in patients with with moderate to severe ileal CD that fail to
respond to steroids or immunomodulators.
Eshuis EJ, et al. BMC Surg 2008
Impact of thiopurines on CD’s natural history
Retrospective study in a hospital cohort of 557 CD patients (CD diagnosis from 1978
to 2002). Assessment of therapeutic requirements and development of disease
complications.
Cohort
1978‐82
Cohort
1983‐87
Cohort
1988‐92
Cohort
1992‐97
Cohort
1998‐2002
Log rank
P value
Patients on IMS
0.04
(.02‐.08)
0.14
(.10‐.18)
0.27
(.23‐.32)
0.45
(.40‐.50)
0.63
(.49‐.76)
<0.00001
Intestinal resection
0.36
(.29‐.48)
0.30
(.25‐.35)
0.32
(.28‐.37)
0.31
(.27‐.36)
0.36
(.22‐.53)
0.528
Large intestinal resection
0.29
(.23‐.35)
0.22
(.18‐.27)
0.19
(.16‐.23)
0.15
(.11‐.19)
0.12
(.05‐.28)
<0.00001
Cumulative probability (95% CI)
Cosnes J, et al. Gut 2005
Drug efficacy depends on timing
RCT in active CD comparing prednisolone 1 mg/kg/day (tapering dose to 12 weeks)
associated to PLACEBO or AZATHIOPRINE 2.5 mg/kg/day for 15 mo (n=63).
Patients stratified regarding ileal, colonic or ileocolonic disease.
Median time since diagnosis at inclusion: AZA 2.6 years (0.1‐19), PBO 3.7 years (0.1‐18.7)
months
Candy S, et al. Gut 1995
Drug efficacy depends on timing
RCT comparing scheduled IFX infusions or episodic (on demand) IFX in active
CD (ACCENT‐1 trial).
Median disease duration at inclusion: 8 years (3.6‐15).
Rutgeerts P, et al. Gastroenterology 2004
How many ‘top‐down’ strategies are there?
surgery
biologicals
thiopurines / MTX
prednisone / budesonide
5‐ASA
antibiotics
efficacy in inducing mucosal healing
Mucosal healing after
INDUCTION therapies
STEROIDS
(week 7)
n=70
Landi B.
Gastro 1992
ENTERAL
NUTRITION
(week 8)
n=25
Fell JME. APT
2000
INFLIXIMAB
(3 infusions)
(week 10)
Rutgeerts P.
GIE 2006
efficacy in maintaining (and inducing?)
mucosal healing
Mucosal healing during MAINTENANCE therapies
Late CD
IFX
scheduled1
1 Rutgeerts
AZA2
Early CD
AZA
postop.3
Step‐up4 Top –down4
(PDN+/‐AZA) (IFX+AZA)
P, et al. ACCENT‐1 substudy, 1 year (n=26). Gastrointest Endosc 2006
2 D’Haens G, et al. 2 years (n=20). Gastrointest Endosc 1999
3 D’Haens G, et al. 1 year (n=41). Gastroenterology 2008
4 D’Haens G, et al. SUTD substudy 2 years (SU, n=23; TD, n=26). Lancet 2008
Prognostic value of mucosal healing in new onset CD
Follow‐up (4 years) of 42 CD patients included in the SUTD study.
Patients remaining in clinical
remission after 4 years
P = 0,004
15/22
7/19
Baert F, et al. DDW 2008