Docstoc

Phase II trial design

Document Sample
Phase II trial design Powered By Docstoc
					        Evolution in Phase II oncology trial design:
            from exploratory to hypothesis testing back to
                        exploratory approach



                        Honoring Ed Gehan
                           St. Louis, MO
                           May 19, 2008


                        Young Jack Lee, Ph.D.
                     LSK Global PS, Seoul, Korea
LSK Global PS
www.lskglobal.com
                                                 LSK Global PS
Evolution in approaches to phase II trials




1960 to1980       Exploratory approach with minimum safety data
1980 to present   Hypothesis driven approach with safety and
                  historical therapeutic data
Future            Exploratory approach with safety, therapeutic,
                  and enrichment data




                                                        www.lskglobal.com
 Principles governing phase II trial designs         LSK Global PS



Acceptance principle:
In phase II studies, the “false positive error” of sending an
ineffective agent on to further study is much less important than
the “false negative error” of concluding that an effective agent is
not efficacious. If the latter error is made, the agent may never
be studied again. When a false positive error is made, further
studies should delineate the place of the agent in the
armamentarium of treatments available for the particular type of
cancer. Gehan (1986)


Rejection principle:
a represents the probability of failing to reject a treatment with
response probability <p0 . This is a less serious error from a drug
discovery viewpoint, but it is serious from a cost perspective
since it leads to unnecessary follow-up trials. Simon (1989)



                                                            www.lskglobal.com
Phase II trial designs                              LSK Global PS



In hypothesis test setting, relative size between a and b will
determine between a rejection design and an acceptance design.

1. Acceptance design
Gehan (1961): Accept the drug for further evaluation if
observed response rate is >a/N. This design is a a >b design
against an ineffective drug.

2. Rejection design
Simon (1989), SWOG (2002): Reject the drug from further
evaluation with a<b requirement if observed response rate is
<r/N.

3. Lee et al (1979) design is a general a>b design with a
statistical grey zone.



                                                            www.lskglobal.com
 Simon optimal two stage design                                    LSK Global PS




     a, b           0.05, 0.1           0.1, 0.1        0.2, 0.1        0.3, 0.1
  p0 vs pA    1st st     2nd st   1st st    2nd st   1st st   2nd st 1st st    2nd st
0.05 vs 0.2   1/21       4/41     0/12      3/37     0/12     2/29   0/12      1/21
0.05 vs 0.3   0/9        2/17     0/7       2/21     0/7      1/15   0/7       1/15
0.1 vs 0.3    2/18       6/35     1/12      5/35     1/13     3/22   0/8       2/18
0.2 vs 0.4    4/19       15/54 3/17         10/37    4/19     6/25   2/13      4/19




                                                                           www.lskglobal.com
Dr. Emil Frei, Harvard Medical School               LSK Global PS
03-10-1997




 Pre-NDA Phase II studies that are positive and confirmed
 should be sufficient to support an NDA. Phase IV trials
 conducted after NDA approval are a more desirable
 environment for exploring exactly how a new agent compares
 with standard treatment.
 Phase II trials should be designed to maximize the
 possibility of a positive result, since false positives will be
 discovered in extended trials, while false negatives likely
 will result in the agent being dropped from further study.
 Phase I studies should be designed with therapeutic as well
 as toxicological intent …




                                                           www.lskglobal.com
 Risks and Benefits of Phase I Oncology                 LSK Global PS
 Trials, 1991 through 2002 (NEJM, 5-Mar-2005)



  Overall response rate (460 trials, 10,402 patients)     10.6%
  Single agent response rate (92 trials)                   4.4%
  Combination therapy response rate (213 trials)          17.8%

  Rate of death due to toxic events                        0.49%
  At least one episode of grade 4 toxic events           14.3%


Rate of death due to toxic events in 1991-1995                    1.1%
Rate of death due to toxic events in 1998-2002                    0.06%
Rate of death due to toxic events in 1991-2002 (213 trials,       0.54%
6474 pts)
November 3, 2004 Journal of the American Medical Association



                                                              www.lskglobal.com
                                                 LSK Global PS




Cetuximab monotherapy and cetuximab plus irinotecan
in irinotecan-refractory metastatic colorectal cancer.


This is a randomized phase II trial with 228 to the combination
therapy and 111 to the monotherapy.


Cetuximab plus irinotecan: 22.9 % [95 % CI (17.5, 29.1)]
Cetuximab monotherapy: 10.8 % [95 % CI (5.7, 18.1)]
New England Journal of Medicine 2004 July 22




                                                       www.lskglobal.com
Understanding the Approval Process                   LSK Global PS
for New Cancer Treatments (NCI)




 A phase II trial asks: does the treatment do what it's
 




 supposed to do, and how well does it work?


 In phase II trials, people with specific types of cancer or with
 certain cancer risk factors are given the medication to see
 whether or not it has a beneficial effect. Each phase II study
 usually focuses on 30-50 patients. Phase II bridges the
 information gap between "is the treatment safe?" to "will it
 actually work in a specific situation?"
 People enrolling in phase II trials may find that beneficial
 effects, when they occur, may still be mixed with side effects
 and complications that the researchers did not expect. Only
 about 33 percent of the drugs tested in phase II trials are found
 safe and effective enough to go on to phase III.



                                                            www.lskglobal.com
 Diverse objectives of Phase II trials              LSK Global PS




a. when testing new agents that have just completed Phase I
trials, to confirm that the dose and schedule chosen can be safely
administered in subsequent Phase II trials;
b. to determine the antitumor activity of existing antitumor
agents that can be administered in significantly higher doses
when used with colony stimulating factors or other factors that
modulate toxicity or antitumor activity;
c. to determine the antitumor activity of combinations of
antitumor agents and modalities;




 PHASE II TRIALS OF NEW ANTI-CANCER AGENTS
 NIH GUIDE, Volume 22, Number 5, February 5, 1993
 RFA: CA-93-09
                                                            www.lskglobal.com
 Diverse objectives of Phase II trials              LSK Global PS




d. to determine the spectrum of antitumor activity for new
agents in selected human cancers.
e. to gain further insight into the pharmacokinetics and
metabolism of the therapeutic agent, its mechanisms of action
and/or toxicity and identification of the particular patient
population most likely to benefit from its effects through the
performance of parallel biological studies.




 PHASE II TRIALS OF NEW ANTI-CANCER AGENTS
 NIH GUIDE, Volume 22, Number 5, February 5, 1993
 RFA: CA-93-09
                                                             www.lskglobal.com
   2006 Conference on Adaptive Trial Design,          LSK Global PS
               Washington, DC



 Scott Gottlieb, MD, Food and Drug Administration July 10, 2006



    Typically, decisions such as how to sample during an
    experiment are made and fixed in advance. In a classical clinical
    trial, patients are allocated to one of two different treatment
    options with half being assigned to each therapy. At the end of
    the experiment a decision is made as to which treatment is more
    effective.

In contrast, in an adaptive clinical trial, patient outcomes can be
used as they become available to adjust the allocation of future
patients or some other aspect of the study design. This allows
researchers to improve expected patient outcomes during the
experiment, while still being able to reach good statistical
decisions in a timely fashion.


                                                             www.lskglobal.com
 2006 Conference on Adaptive Trial Design,          LSK Global PS
             Washington, DC




A second type of adaptive trial design involves ongoing
assessment of the sample size, to avoid under- or over-allotment
of patients. For example, if the statistical power of a trial is
based upon a particular variable and an estimate of its variance,
it is easy to see how an increase or decrease in the variability of
the sample could affect the power. By continuously monitoring
such a critical factor, it is possible to adjust the sample size of a
trial for the power that is desired.

There is much interest, therefore, in being able to carry out
interim assessments of long running trials to ensure that the
design is still appropriate to meet its needs or that its safety and
efficacy data do not indicate that the trial should be modified or
even stopped. This involves much more than the now traditional
concept of predetermined stopping point criteria.




                                                           www.lskglobal.com
 2006 Conference on Adaptive Trial Design,          LSK Global PS
             Washington, DC



Adaptive designs could have other benefits. They can help us fill
in the frustrating white spaces between phases, enabling
seamless designs that allow learning to be more iterative and
less method-limited. That allow continuous discovery that isn’t
defined by phases but rather by what we learn as we go. An
adaptive design can also be more effective than standard designs
at identifying 'the right dose'. And it usually identifies the right
dose with a smaller sample size. Another advantage is that many
more doses can be considered in an adaptive design, even though
some may be little used or even never used.

The guidance documents we are developing include one to help
guide sponsors on how to look at multiple endpoints in the same
trial. This guidance document is currently being drafted and we
hope to be able to discuss that work as soon as January.
Another guidance document that we are also working on now
deals with enrichment designs, designs that can help increase
the power of a trial to detect a treatment effect, potentially with
fewer subjects.

                                                           www.lskglobal.com
                                                                 LSK Global PS
  Counter examples for today’s phase II design


 Unreliable response measurement

Endpoint metastic melanoma       Genasense + DTIC   DTIC alone    Hazard     p-value
                                 (N=386)            (N=385)       Ratio
Survival                         274 days           238 days      0.89       0.18
PFS                              61 days            48 days       0.75       0.006
Response Rate by Investigators   11.7%              6.8%                     0.019
Response Rated Confirmed by      6.7%               3.6%                     0.056
Independent Review Committee



 Ignored study design

Because the drug (Gleevec) appeared so promising early in the
trial, the study -- originally planned to include only 36 patients --
was increased to 150 patients to allow more people access to the
drug.


                                                                         www.lskglobal.com
                                                        LSK Global PS
Counter examples for today’s phase II design



  Poor response rate of study compounds, particularly single
 compound, making recruitment difficult.



  Non-robustness of design: sensitivity to assumed
 response rates under null and alternative hypotheses


  Simon optimal two-stage design for =0.05 and  =0.1
    pA              0.19                0.20                0.21
    p0     1st st     2nd st   1st st     2nd st   1st st     2nd st
    0.04   0/15        3/36    0/12        3/37    0/11        3/36
    0.05   1/20        5/57    1/21        4/41    1/18        4/44
    0.06   1/22        6/58    1/20        6/57    1/20        5/46


                                                                   www.lskglobal.com
 Unreal two-stage rejection design                             LSK Global PS



 Finding an agent of 10% effectiveness could also be important
 for patients with lymphoma, if the 10% response rate was
 additive to the response rate for existing combination
 treatments. Gehan (1986)

    a,b           0.05, 0.1          0.1, 0.1         0.2, 0.1          0.3, 0.1
 p0 vs pA     1st st   2nd st    1st st   2nd st   1st st   2nd st   1st st   2nd st
0.1 vs 0.2    5/47     18/130 4/41        13/99    3/34     8/68     1/23     5/46
0.15 vs 0.25 9/57      31/162 8/52        23/125   6/43     14/81 3/31        9/55
0.2 vs 0.3    15/71    45/184 12/59 34/144         10/52 21/93 5/34           14/65


 Randomized phase II in metastatic colorectal cancer
 Cetuximab plus irinotecan: 22.9 % [95 % CI (17.5, 29.1)] N=228
 Cetuximab monotherapy: 10.8 % [95 % CI (5.7, 18.1)] N=111
 New England Journal of Medicine 2004 July 22

                                                                        www.lskglobal.com
                                                                            LSK Global PS
     Helpless two stage rejection rule



Finding an agent of 10% effectiveness could be of crucial
importance for patients with advanced lung cancer. Gehan (1986)

What should be the study design for finding an agent of 10%
effectiveness?

     a ,b              0.05, 0.1             0.1, 0.1             0.2, 0.1            0.3, 0.1
 p0 vs pA         1st st      2nd st    1st st    2nd st      1st st    2nd st    1st st    2nd st
0.05 vs 0.1                             5/101 13/195          3/78      8/131 2/56          6/115
0.02 vs 0.1       0/25        1/44      0/26      2/59        0/25      1/44      0/25      1/44
0.01 vs 0.1       0/24        2/65      0/24      1/47        0/22      0/23      0/22      0/23

Iressa was approved in May 2003 on the basis of a surrogate endpoint which was approximately a
10% tumor response rate (tumor shrinkage) in clinical studies in patients with advanced lung cancer.

Is any rule better than the 1/22 rule of Gehan (1961) for this case?
                                                                                      www.lskglobal.com
Difficult recruitment                              LSK Global PS




For phase I and II cancer clinical trials, recruiting is a
particular challenge, as researchers strive to accrue cancer
patients into trials that are designed to answer scientific
questions and improve treatment overall, but are not
designed specifically to try to make the participants well.
WESTAT: “Enhancing Recruitment to Early Phase Cancer
Clinical Trials: Literature Review”, a report submitted to NCI,
January 5,2004




                                                          www.lskglobal.com
Conclusions/Discussion                             LSK Global PS


The assumption pA-p0>0.2 is not realistic. The assumption is to
make the sample size artificially within 30-50. A more realistic
assumption is pA-p0<0.1.
FPP>FNP designs may be preferred, namely acceptance rule
rather than rejection rule. This design will make the sample
size within 30-50 for a 10% additive effectiveness.
Design driven by the # of responses rather than total N should
be considered. This makes the design exploratory and flexible.
Cumbersome and unrealistic “white space” between stages,
which is usually ignored, can also be removed.
Most new cancer drugs are added to the existing regimen and
tested in late phases. Many phase I and early phase II trials are
single-agent studies, and face difficulty recruiting patients. Is
it ethical to continue the practice?
Having to specify null and alternative hypotheses and
determining the sample size for phase II trials in emerging part
of the world is a really onerous nuisance. Why do we need
hypothesis test setting for phase II trials?
                                                          www.lskglobal.com
                                                     LSK Global PS
Phase II trial design today and future




  Today                             Future
  Estimating anti-tumor activity Diverse study objectives
  Safety data                       Safety, therapeutic, and
                                    enrichment data
  Rejection design                  Acceptance design
  Inflexible decision rule          Flexible decision process
  Fixed design: sample size         Flexible adaptive design: # of
  driven                            responses driven
  One primary endpoint              Multiple endpoints
  White space between stages        Seamless process
  Phase II to phase III             Phase II to market
  Single or multi-institutional     Multi-national


                                                           www.lskglobal.com
                                                 LSK Global PS
Time to pick Ed’s brains




1960 to1980       Exploratory approach with minimum safety and
                  few therapeutic data
1980 to present   Hypothesis driven approach with safety and
                  historical therapeutic data
Future            Exploratory approach with safety, therapeutic,
                  and enrichment data




                                                        www.lskglobal.com
                                 LSK Global PS




Thank you, and I am glad to meet you all.




                                      www.lskglobal.com

				
DOCUMENT INFO
Shared By:
Categories:
Tags:
Stats:
views:4
posted:10/28/2011
language:English
pages:23
xiaohuicaicai xiaohuicaicai
About