Endpoints in clinical studies

1
Endpoints in clinical studies

• Mark Conaway
• Div of Biostatistics and Epidemiology

2
Outline

• Background why discuss endpoints?
• Common theme to observational studies and RCTs
• Importance to clinical research
• Multiple endpoints
• Surrogate endpoints

3
Endpoints

• Quantitative measurements required by or implied
  by the objectives of the study/trial
• Often see distinction between
• hard and soft endpoints or
• objective and subjective endpoints

4
Prefer hard endpoints

• hard endpoints clinical landmarks that are
• well-defined in study protocol
• definitive with respect to disease process
• not subjective
• Examples
• death,
• time to disease progression/relapse,
• some laboratory measurements

5
Soft endpoints

• Soft endpoints
• not directly related to disease process or
• require subjective assessment by
  patient/physician
• Examples
• Quality of life questionnaires
• Symptom questionnaires

6
Not all endpoints can be classified

• Some endpoints are useful and reliable but
  require some subjectivity
• Examples
• Pathology
• Imaging

7
Why is a discussion of endpoints important?

• Original intent of this lecture was as a
  follow-up to
• the observational studies lecture (Dr. Bovbjerg)
  and
• the randomized clinical trial lecture (Dr.
  Petroni)
• Both are attempts to apply rigorous scientific
  principles to study benefits of therapies, so why
  do they disagree so often?

8
Disagree

• Findings suggested in observational studies are
  not the same as those found in the RCT
• Usually that means an effect suggested by
  observational studies is not supported in the
  RCT.

9
Many reasons for this..

• Different patient populations?
• Potential confounders not adequately controlled
  for in the observational studies?
• Design of RCT?
• Choice of endpoints?
• Hobart et al, 2007 raises issue that potentially
  useful therapies in neurology have been discarded
  because of the RCT endpoints used.

10
Example association between periodontal disease
and preterm birth

• Michalowicz et al Treatment of Periodontal
  Disease and the Risk of Preterm Birth (NEJM, Nov
  2, 2006)
• Multi-center RCT of 820 patients, half assigned
  to receive periodontal therapy during pregnancy,
  half to received usual dental care.

11
Supporting evidence for the trial epidemiological

• Offenbacher et al (2006) OCAP study
• 1020 pregnant women, assessed periodontal health
  at enrollment (lt 26 wks GA) and postpartum.
• Results rates of preterm births (lt 37 wks)
• Healthy 11.2 , Mod-Severe dx 28.6
• Rates of very pre-term births (lt 32 wks)
• Healthy 1.8, Mod-Severe dx 6.4

12
Supporting evidence for the trial epidemiological

• Goepfert et al (2004) case control study
• cases 59 women with spontaneous delivery birth
  at lt 32 weeks
• Controls 44 women with early indicated births
  of lt 32 weeks and 36 women with term births
• Results
• Cases more likely to have periodontal disease

13
Supporting evidence for the trial epidemiological

• Boggess et al (2003)
• Single cohort of 763 births
• Studied rate of preeclampsia (39 women)
• Results Periodontal disease associated with risk
  of pre-eclampsia
• Many other papers looking at birth weight, Apgar
  scores,....

14
Other evidence

• Small intervention trials
• Animal studies to assess specific types of
  bacteria
• Overall, a good deal of evidence to support the
  notion that reducing periodontal disease might
  reduce rates of problems

15
Multi-center RCT (Michalowicz et al)

• Results Gestational age at delivery
• No difference between the groups
• Why?
• Different patient populations?
• Treatment effective enough?
• Was this the right endpoint?

16
What other endpoints?

• Rate of pre-term birth
• lt 37 weeks ? Or lt 35 weeks Or lt 32 weeks?
• Birth weight?
• Gestational age at delivery?
• Stillbirths?
• Apgar scores?
• Admission to NICU?

17
The study also considered other possible endpoints

• Rate or pre-term birth (lt 37 weeks)
• Birth weight
• Proportion of infants who are small for
  gestational age
• Apgar scores
• Admissions to NICU
• Stillbirths (5 in treated group, 14 in control
  group, p 0.08)

No difference
18
One source of discrepancy

• Can see a situation where the observational
  studies put forth the best endpoint
• Is this a good idea?
• Clinical trials focus on one particular endpoint
  as primary
• Why?

19
In principle

• Have chosen an endpoint
• Clinical relevant
• Have an appropriate sample size
• to have a type I error rate of 5
• sufficient power for a clinically meaningful
  difference

20
In practice

• Rare that trials use a single endpoint
• Endpoints
• cover clinical events
• symptoms
• physiologic measures
• side effects
• quality of life

21
Whats the problem?

• If test each endpoint at the 5 level
• overall chance of finding at least one endpoint
  where there is a significant difference is larger
  than 5, even if the treatments are identical
• Prone to distorted reporting (i.e. pick most
  significant)
• Good reference Pocock (1997) Controlled Clinical
  Trials, p 530-545

22
Whats the problem?

• If you lock in on a single endpoint, in either
  the RCT or observational studies
• Could miss important therapies
• If you allow yourself to look at many endpoints
• Could wind up doing RCTs based on spurious
  findings or
• Recommending treatments that arent truly
  effective

23
What to do?

• Have a pre-defined strategy
• Some advocate
• all results pre-written, with results filled in
  as trial concludes
• Alternative view
• need to be flexible
• need to allow for unexpected findings
• but recognize potential for problems type I
  error rate is not 5

24
Delineate primary and secondary outcomes

• Many advocate having a single primary endpoint
• drives sample size calculations
• test based on this endpoint has a 5 type I error
  rate
• All other endpoints are secondary

25
Example Michalowicz et al (NEJM, Nov 2, 2006)

• Primary
• Gestational age at end of pregnancy
• Secondary
• Pre-term births (lt 37 weeks), birth weight,
  proportion of infants who are small for
  gestational age, Apgar scores, admissions to NICU

26
Delineate primary and secondary outcomes

• Can be hard to adhere to in practice
• For example, what if primary outcome is not
  different among groups, but all secondary
  outcomes are?

27
Bonferroni procedure

• If you have k endpoints
• Multiply observed p-value by number of endpoints
• For example, with k 8, convert an observed
  p-value of 0.01 to 0.08
• Ensures that if Ho is true for all endpoints,
  probability of rejecting Ho for at least one
  endpoint is less than or equal to ?

28
Limitations forBonferroni procedure

• Endpoints tends to be correlated, so this is
  conservative
• probability of type I error is much smaller than
  ?
• Treats all outcomes as equal in importance
• Can lead to illogical results
• Trial with p-values 0.01, 0.75,0.75,0.75,0.75
  significant
• Trial with p-values 0.02, 0.02, 0.02, 0.02, 0.02
  is not significant

29
Limitations forBonferroni procedure

• Procedure reduces power to detect real
  differences in specific outcomes, if they exist
• Protect type I error at expense of power
• Difficult to apply strictly in many cases

30
So whats the answer?One persons opinion...

• Selection of a primary outcome is important
• Need to allow for surprises
• Full disclosure of endpoints, instead of selected
  endpoints, can alleviate a lot of the problems
• Adjust p-values?
• Whats the goal of the study?

31
Surrogate endpoints

• Hesitate to use the term
• Has a specific technical definition
• Issue
• Quicker, less expensive, less clinically relevant
  endpoint or
• More expensive, clinically definitive endpoint?

32
Example

• Treatment for osteoporosis
• Endpoint
• Bone density via DEXA?
• Fracture?
• If fracture, how would this be ascertained?

33
Example

• Choice is, for same amount of resources
• more patients with less clinically relevant
  outcome (bone density)
• Fewer patients with more clinically relevant
  outcome (fracture)
• Frequently see the quick endpoint in earlier
  stage trials.

34
Summary

• Choice of endpoints is crucial to the success of
  the study either RCT or observational
• Issues about
• Which one?
• How many?
• Primary vs secondary?