Stratification Carryover From Session 1

1
Session 2

• Stratification (Carry-over From Session 1)
• Site selection
• Concomitant therapies
• Required
• Allowed
• Prohibited
• Washout and wash in periods
• Change in medication regimen during the study
• Special considerations about antibiotics

2
To Stratify or Not to Stratify?

• Possible considerations for stratified
  randomization
• Concomitant therapies
• Disease severity
• Clinical site

3
Stratification

• In principle, randomization should lead to
  balanced distribution of important prognostic
  variables
• In practice, imbalances can and do occur
• Notable imbalances may affect study credibility,
  even though analyses are adjusted
• Need to limit stratification to only the
  strongest prognostic variables using too many
  variables will defeat purpose
• Try to stratify by site if sample size at each
  site will be large enough

4
To Stratify or Not to Stratify?

• What would you do?

5
To Stratify or Not to Stratify?

• What would you do?
• What we did
• Stratify on clinical center only
• Each center with their own randomization order
• Minimize any effect of variation in execution of
  the protocol

6
Choosing the proper sites will make your life as
PI and/or Project Manager a pleasure. Choosing
the wrong sites can ruin even the best of ideas.
7
Site Selection

• Grantsmanship vs. Real World
• Name value
• Track record of recruitment
• Feasibility of coordinating subcontracts prior to
  grant application
• Scientific input
• Investigator initiated is different from industry
  sponsored trials
• Additional skills special assay, etc.

8
Does Site Screening Require An Onsite Visit?

• Pictures tell more than words
• Drug storage
• Patient encounters
• Record storage
• Meet all the people
• Not easy to see the problems from an informal
  visit

9
Will The Site Perform Well?

• Prior experience
• Reputation
• Population-size
• Level of enthusiasm
• Scientific
• Financial
• Competing interests
• Track record of recruitment

10
Screening Questions

• See OHR website for sample form.

11
(No Transcript)
12
(No Transcript)
13
How to Recruit Sites

• Cold letter or cold call
• Referral from other sites
• Web-based advertising

14
PredictorsAnecdotes From My Experience

• Good predictors of site performance
• Enthusiasm
• Recommendations of other PIs
• Poor predictor of site performance
• First or last author on lots of industry
  sponsored trials

15
Some (Conventional) Wisdom About Estimates of
Recruitment

• Whatever number a site projects for the number
  they can enrolldivide by 10!
• Consider whether there are or will be competing
  studies ongoing, and how priorities will be set
• Good recruitment in single-site studies does not
  necessarily predict good recruitment in a
  multicenter trial

16
Session 2

• Stratification (Carry-over From Session 1)
• Site selection
• Concomitant therapies
• Required
• Allowed
• Prohibited
• Washout and wash in periods
• Change in medication regimen during the study
• Special considerations about antibiotics

17
Ulcerative Colitis Mild to Moderate
Responseadequate
Responseinadequate
Response adequate
Considerincreased dose
Response inadequate
Oral 5-ASA
Responseinadequate
Responseadequate
Response inadequate
Sands B. 2004
18
Ulcerative Colitis Moderate to Severe
Inadequate response
Inadequate response
Adequate response
Response
No Response
Unsuccessful
Failure
Sands B. 2004
19
Required Concomitant Therapies

• None
• 1st line therapies
• 1st and 2nd line therapies

20
Ethical Considerations

• How dangerous is it to not treat
• Can you have equipoise if you are not treating
  some patients
• Are second line treatments clearly effective
• How safe are second line treatments
• If prior intolerance, can you require rechallenge

21
Practical Considerations

• Access to patients who have not been treated
• Will patients agree to enroll in a trial with an
  arm that is no therapy?
• If prior intolerance, will not likely be willing
  to rechallenge

22
Other Considerations

• Are patients and physicians OK with multiple
  therapies?
• If therapies are likely to be used together, best
  to study together
• Allow concomitant treatment but specify which one
  it has to be?
• If therapies are likely to be added if patient
  remains symptomatic, may want to build a strategy
  into the protocol if possible

23
Required Concomitant Therapies

• What would you do?

24
Required Concomitant Therapies

• What would you do?
• What we did
• Minimum of 4 weeks of 5-ASA at 2gm/day or higher
  dose, or
• Intolerant to 5-ASA

25
Prohibited Therapies

• Why prohibit any concomitant therapy
• Marker of disease too severe for inclusion
• Experimental therapies
• How to define
• Known or anticipated interaction

26
Permitting Other Experimental Therapies

• Practice greatly resisted by drug makers
• Complicates assessments of adverse effects
• If therapy aimed at same symptoms, also
  complicates efficacy assessment
• Sometimes warranted
• Early days of AIDS few drugs, almost all
  experimental
• Unreasonable to ask subject in trial of
  antiretroviral drug to not participate in trials
  of drugs to treat or prevent opportunistic
  infections

27
Prohibited Therapies

• What we did
• Cyclosporine and tacrolimus
• Too severe of disease
• Experimental therapies
• Anti-TNF therapies
• Methotrexate
• Clinical trials

28
Allowed Therapies

• 2nd or 3rd line therapies
• Pro
• Facilitates enrollment
• Increases generalizability
• Con
• Greater chance for drug interaction
• Potential for imbalance of the second line
  therapies

29
Drugs We Had To Consider

• Oral steroids
• Toxicity with long term therapy
• Dose dependent toxicity
• Flare of disease with tapering could blur the
  signal of efficacy if those not on active therapy
  are less likely to try to taper
• One goal of a new therapy could be steroid
  tapering
• 4 of 8 patients tapered off steroids in pilot
  study

30
Drugs We Had To Consider

• Azathioprine / 6-MP
• Slow time of onset
• Unclear duration of effect after discontinuation

31
Drugs We Had To Consider

• Rectally administered therapies
• Improves symptom control by treating the rectum
• Increase symptoms with stopping
• Impediment to recruitment
• Could change histology of distal colon where
  biopsies would be obtained.

32
Allowed Therapies

• What would you do?

33
Allowed Therapies

• What would you do?
• What we did
• Steroids at maximum dose of prednisone 20mg/kg,
  no tapering
• Azathioprine / 6MP allowed
• Rectal therapies allowed but required to continue
  at current dose throughout
• This was an amendment to facilitate recruitment

34
Washout and Wash In Periods

• How long does it take to have a stable effect of
  starting, stopping, or changing a dose?
• Clinical data when available
• PK data less reliable
• Practical issue
• How long will patients and physicians wait before
  starting a new therapy?
• Can impact recruitment

35
Note on Washout Periods

• Difficulty in defining time after which there
  will be no carryover effect
• Leads to great caution in planning crossover
  trials as definitive approach to establishing
  drug efficacy
• In some medical areas (e.g., schizophrenia),
  patients undergoing washout must be very
  carefully monitored

36
Final Inclusion Criteria

• Disease Activity Index score 4 and 10
• 5-ASA
• 2 gm/day for at least 4 weeks or intolerant
• Steroids
• Stable dose for 4 weeks prior to entry
• 20mg prednisone or equivalent
• 6MP or azathioprine
• 4 months and stable dose for 2 months prior to
  entry
• Rectal therapy
• Stable dose for 2 weeks prior to entry

37
Unanticipated Issues

• Medication was not working so it was stopped. How
  long does the patient need to be off of this?
• If not effective, is there a washout effect?
• Recruitment considerations
• Generalizability to real world settings
• What would you do?

38
Issue of Prior Medication

• Even if medication didnt seem to be effective,
  might have residual action that could interfere
  (or synergize!) with effect of treatment you are
  studying
• PK or other information on likelihood of extended
  effects?
• Any basis for defining a washout period?

39
Discontinued Medications

• We did not require a washout period

40
Unanticipated Issue

• Antibiotics
• Weak data that have any affect on UC
• Occasionally tried
• Used for lots of other conditions
• Do you need an antibiotic washout period?
• What if subject requires short-term antibiotics
  during the study for treatment of infection?
• Cant very well prohibit this

41
Antibiotics

• We had no rules about antibiotics
• Very few patients were exposed

42
Session 3 External Group Contacts

• NIH
• IRB
• FDA
• Industry
• DSMB