Study Monitoring

1
Study Monitoring DSMBs
April 18, 2007

• Marshall J. Glesby, MD, PhD
• Cornell HIV Clinical Trials Unit
• Division of International Medicine and Infectious
  Diseases

2
Purpose of Study Monitoring

• Ongoing evaluation of risk benefit
• Ethical responsibility to stop a trial if
  intervention is harmful or clearly beneficial
• May identify need to collect additional data
• May uncover logistical problems or poor data
  quality that need to be addressed

3
Reasons for Stopping Early --1

• Study tx found to be convincingly better or worse
  than standard tx
• Toxicity too severe to continue tx
• Data of poor quality
• Accrual too slow to complete in timely fashion
• Definitive information about the tx becomes
  available making the study unnecessary/unethical

Piantadosi S, Clinical Trials a Methodologic
Persrpective, 2nd ed. 2005
4
Reasons for Stopping Early --2

• Finding a statistically significant difference by
  the end of the study is improbable
• Scientific questions are no longer important due
  to other medical advances
• Adherence to tx is unacceptably poor, preventing
  an answer to the basic question
• Study integrity undermined by fraud/misconduct

Piantadosi S, Clinical Trials a Methodologic
Persrpective, 2nd ed. 2005
5
Tension Regarding Decision to Stop

• Pressure to terminate trial at earliest possible
  moment
• Minimize pts receiving inferior tx
• Minimize size, duration of trial (sponsor)
• Disseminate results early (investigator)
• Benefits of continuing the trial as long as
  possible
• Increase precision of tx effect estimate
• Obtain sufficient power to account for effects of
  prognostic factors and be able to perform
  pre-specified subgroup analyses
• Be able to analyze secondary endpoints

6
Ethical Aspects

• Collective good (what well learn from the trial)
  vs. individual good (risk/benefit of individual
  participants)
• Ethical mandate to maintain integrity of trial
  until it provides a standard of evidence
  appropriate for the setting

7
Possible Recommendations

• Stop the study
• Stop further enrollment
• Stop a study arm
• Modify the design
• Disseminate the results?

8
DSMB Exercise
9
Phase II Trial of Vicriviroc

• Investigational CCR5 receptor antagonist being
  studied in 118 treatment-experienced HIV
  patients by the AIDS Clinical Trials Group
  (protocol chair R Gulick)
• Subjects who enrolled were heavily treatment
  experienced with advanced HIV disease and low CD4
  counts
• Randomized to 1 of 3 doses of vicriviroc vs.
  placebo

Schering Plough News Release March 3, 2006
10

• ACTG Study Monitoring Committee informed the ACTG
  and Schering Plough that 5 cases of malignancy
  were observed in subjects treated with vicriviroc
• 4 pts with lymphoma, 1 with gastric
  adenocarcinoma

Schering Plough News Release March 3, 2006
11

• What other information would you want?
• What would you recommend?

12
Outcome

• a causal association between vicriviroc and the
  lymphoma cases could not be established at this
  time
• Of 4 lymphoma cases, 2 of non-Hodgkins cases had
  prior histories of Hodgkins disease
• ... since vicriviroc, at the 2 highest doses
  tested, showed evidence of virologic activity
  and CD4 count increases, the trial will
  continue.
• Subjects investigators informed of results
• Subjects unblinded so they and their physicians
  may make a more informed decision about their tx
  options
• Long-term (5 yrs) follow-up introduced