Outcome Measures for Clinical Trials in IBD

							Outcome Measures for Clinical Trials in IBD

Introduction

Randomized Clinical trials (RCTs) serve multiple purposes in the development of
new treatments for IBD.          Most commonly these studies provide benchmark
information for clinical practice. While this is universally true in clinical medicine
it is especially valid in IBD where high placebo response rates are observed and
relatively "soft" measures of response are commonly applied. However RCTs
can also provide new insights into disease mechanisms. A recent example is
the observation that administration of an antibody to tumor necrosis factor alpha
results in clinical benefit in Crohn's disease. This finding strongly suggests that
TNF plays an important role in pathogenesis. Moreover, clinical trials also offer
an ideal opportunity to collect biologic samples for basic research and correlate
findings with clinical outcomes. An obvious application of this possibility is
identification of mechanisms of action of drugs. Generally speaking,only limited
use of this opportunity has been made by IBD researchers.


A wide range of outcomes can be assessed in RCTs depending on the specific
type of study. For the purposes of this discussion the following classification
scheme, which is widely employed in           pharmaceutical development, will be
used.    Phase I trials      include initial safety, tolerability and pharmacokinetic
investigations in normal healthy individuals.     Phase Ib studies are a variant of
this type in which early phase investigations are performed in volunteers with the
disorder of interest rather than healthy people. Phase II studies focus on proof of
concept and identification of the optimum drug dose.            Phase III trials are
definitive comparisons of a new agent against either a placebo or a treatment of
known efficacy.       Phase IV studies are increasingly concerned with either
identification of uncommon adverse events or the "effectiveness" (in distinction to
efficacy) of new drug in the real world of clinical practice.


Outcome Measures for Clinical Trials in IBD                                         1
Feagan, BG
It should also be remembered that the RCT is          the most rigorous means of
determining the clinical utility of diagnostic tests, practice guidelines and care
paths.

An Overview of Outcome Measures

A hierarchy of potential outcomes is displayed in Figure 1.

Laboratory Surrogates

In other diseases laboratory tests have emerged         as surrogates for clinically
important events. For example following publication of the Diabetes Control and
Complication Trial, it was accepted that glycemic control, as measured by the
concentration of glycosylated hemoglobin in the blood was highly correlated with
the development of complications of Type I diabetes. This observation formed
the basis of current development programs in diabetes i.e. a beneficial effect on
glycosylated hemoglobin is now accepted by regulatory agencies as proof of
clinical efficacy. Identification of a valid surrogate measure greatly simplifies the
performance of early (Phase I/ Phase II) drug development since it is inefficient
to perform large-scale studies dose ranging studies using clinically relevant
endpoints. Attempts have been made to develop surrogate markers in ulcerative
colitis and Crohn's disease, notably measurement of acute phase reactants (C-
reactive protein? /orosomucoid) and, more recently, markers such as cytokine
concentrations in serum or stool or products derived from fecal leukocytes.
However no valid surrogate measure is currently accepted for use in dose
ranging studies.      Development of such an outcome would be an important
advance.

Imaging Studies

The next step in the hierarchy is imaging studies. Traditional imaging methods
such as radiology and endoscopy have been used to a limited extent as


Outcome Measures for Clinical Trials in IBD                                        2
Feagan, BG
outcomes in clinical trials. The role of endoscopy bears special comment. While
endoscopic outcomes are attractive conceptually they have not been widely
adopted in Crohn's disease for several reasons. First, reliable and practicable
endoscopic severity indices have not been available.       The most widely used
instrument, the CDEIS, has been validated using           appropriate    methods.
However this instrument is relatively cumbersome to score and there have been
concerns regarding reproducibility. Furthermore, the clinical interpretation of
changes in this scoring system remains unknown.         Second, a relatively poor
correlation exists between endoscopic improvement and resolution of clinical
symptoms with traditional therapy for Crohn's disease. However, our initial
experience with infliximab now supports the existence of such a correlation .
Finally, colonoscopy is invasive and expensive. These are important limitations if
repeated procedures must be performed as is usually the case in a RCT. The
situation is quite different for ulcerative colitis where a good correlation exists
between the modified Baron classification stain and clinical symptoms.
Moreover, sigmoidoscopy can be performed at a low-cost and with minimal
inconvenience/risk to the patient.


Recently, several novel imaging methods have become available including high-
resolution CT scanning, ultrasonography with doppler blood flow measurement,
magnetic resonance imaging, and SPECT and PET scanning.                  With the
exception of ultrasound these modalities all have important limitations, most
notably cost and, in the case of CT, SPECT and PET, exposure to ionizing
radiation. Longituidinal studies correlating these modalities with each other and
with clinically relevant measures such as disease activity and health related
quality of life have not been performed.

Clinical Disease Activity Indices

Clinical disease activity in IBD is measured using composite rating scales. Since
relevant changes in clinical disease activity scores currently are the basis for


Outcome Measures for Clinical Trials in IBD                                      3
Feagan, BG
approval of new drugs by regulatory agencies a thorough understanding of these
measures by investigators and sponsors is vital to both scientific research and
patient care. The classic example is the CDAI, a measure which incorporates
subjective    patient    assessments,         clinician   measurements   and   laboratory
evaluation. A systematic review of the operating properties of this index and
other outcome measures in Crohn's disease have recently been published. An
NIH sponsored workshop will review the topic of clinical outcomes in Crohn’s
disease in January 2003. The CDAI remains the gold standard for assessment
of disease activity in Crohn's disease. Several modifications to this measure
have been proposed (elimination of nonresponsive items, censoring for excess
weight). Assessment of the importance of these proposed modifications and
validation of a modified index is desirable. With respect to ulcerative colitis, a
number of clinical disease activity indices have been developed and utilized.
(Mayo Clinic Score, Sutherland, Rachmilowitz). In general, less is known about
the operating properties of these instruments (validity, responsiveness, reliability)
than the CDAI. A systematic review of this area, perhaps under the aegis of the
NIH, would be helpful.

Health Related Quality of Life Measures

Measurement of health related quality of life (HRQL) in IBD has advanced
steadily over the past decade. HRQOL questionnaires are now routinely utilized
as measures of outcome in RCTs.                      The Inflammatory Bowel Disease
questionnaire (IBDQ), the most widely used instrument has been extensively
used in randomized controlled trials and epidemiological surveys.                  Good
validation data are available for Crohn's disease, however relatively less
information is available for ulcerative colitis. The recent commercialization of the
IBDQ has, in some instances, limited its utilization by investigators. The Rating
Form of Health Concerns is another disease specific measure which focuses on
patient's emotional and psychological response to inflammatory bowel disease.
Recently, the SF- 36 a widely used generic HRQL measure, has been introduced


Outcome Measures for Clinical Trials in IBD                                            4
Feagan, BG
into clinical trials are therapy in IBD. Additional data are required regarding the
operating properties of the SF 36, however clinically relevant changes following
infliximab therapy have been shown . Since a wide range of normative and
disease specific data are available for the SF 36 is particularly useful for
comparisons among different disease.              Utility measures are generic HRQL
assessments which are essential               for cost utility analyses.   Only limited
experience is available with these measures as outcomes in controlled trials.
Surrogate methods of utility estimation (Health Utility Market 3, Euro Qual) are
currently being evaluated.        Economic assessments, namely measurement of
direct and indirect costs in IBD has been a neglected area general in general.
Only recently have attempts been made to collect a relevant economic data
during the course of RCTs. Little is known about the HRQL of specific
subpopulations of patients with IBD (proctitis, fistulizing Crohn's disease,
pouchitis). Development of valid disease specific instruments in the special
populations may be necessary and desirable.

Disease Related Complications, Hospital Admission, Surgery and Death

No clinical trial has utilized any of these measure as a primary measure of
response      since they occur relatively infrequently in the total population of
patients with IBD. Nevertheless these outcomes are highly relevant to patients,
providers and payers. With the development of new biological approaches to the
treatment of IBD the question of whether early treatment is able to change the
natural history of these diseases will become highly relevant. Given that large-
scale trials, including more than 500 patients, are now feasible, it is conceivable
that trials with sufficient statistical power to detect differences in surgical rates
and admission to hospital can be designed.                  Such studies will require
international cooperation and should focus on high-risk subgroups.




Outcome Measures for Clinical Trials in IBD                                           5
Feagan, BG
Some Suggestions for Future Clinical Trials Outcomes Research

Based on the preceding discussion the following are suggestions for future
research activity . No specific prioritization has been given.


1. Development of a valid, sensitive and specific surrogate outcome measure
which could be utilized in the performance of Phase II A. studies. Preliminary
investigations concentrating on cytokine profiling seem promising.
2. Define the relative value of MRI/gadolinium vs. ultrasonography in assessing
disease activity and severity in Crohn's disease.            Can either of these
technologies be incorporated into RCTs as objective measures of response?


3.   An overview of current knowledge regarding the operating properties of
clinical disease activity indices in all inflammatory bowel disease should
performed. Clinical investigators should come to a consensus agreement on
disease definitions and measurement issues.


4. The determinants of the placebo response rate should be investigated in both
ulcerative colitis and Crohn's disease.       Understanding these factors would
improve the efficiency of RCTs.


5.   Small to medium-size RCTs should be performed which are designed to
explore clinical-basic science correlates in well-defined groups of patients.
Techniques such as differential display, gene chip technology and nuclear
resonance spectroscopy should be utilized to identify the mechanisms of action
of corticosteroids, the purine antimetabolites, methotrexate and infliximab.


6. Disease specific quality of life measures should be developed for specific
subtypes of patients. (Pouchitis, perianal fistula, proctosigmoiditis). Utility and
health care resource utilization data should be collected from RCTs in these
specific patient populations.

Outcome Measures for Clinical Trials in IBD                                      6
Feagan, BG
7. Investigators performing large-scale clinical trials should be encouraged to
collect blood samples and possibly tissue which can be analyzed for known
laboratory markers (NOD2, ASCA, ANCA). Identification of specific phenotypes
of disease should be a collaborative goal with a common database.          Again,
standardization of terminology and outcomes would be essential for a successful
large-scale initiative.      Collaboration with "gene hunting" groups would be
essential.
8. The value of therapy drug monitoring for existing treatments, specifically the
purine antimetabolites, should be defined.


9. The potential benefit of early aggressive medical therapy on outcomes such as
hospitalization and requirement for surgery should be defined by large-scale
RCTs in high-risk subgroups of patients.


10. Epidemiological studies should define the rates of severe adverse events
potentially related to immunosuppression (opportunistic infection, severe
bacterial infection, tuberculosis, lymphoma, skin cancer).     Estimates of these
rates should be used to develop expected frequencies of serious adverse events
for active treatment groups in RCTs.          This would be an important tool for
understanding the results of trials of new agents.


11. Existing multicenter clinical trial groups should be encouraged to work more
closely together. The CCFA and CCFC should work together to develop joint
funding mechanisms for RCTs.




Outcome Measures for Clinical Trials in IBD                                     7
Feagan, BG