Response Evaluation: Beyond RECIST

1
Response Evaluation Beyond RECIST

• Elizabeth Eisenhauer MD FRCPC
• ESMO Conference Lugano
• 07 July 2007

2
Outline

• Background
• Why measure response?
• Response criteria in cancer trials
• Key aspects of RECIST
• Implementation issues with RECIST
• Minimum number of lesions
• RECIST in randomized trials
• Imaging with MRI and PET
• Is RECIST applicable in trials of non-cytotoxics?

3
Why Measure Response?

• The word response is used in a number of
  contexts
• To describe outcomes in daily practice (my
  patient is responding to treatment)
• As a surrogate for benefit (e.g. in randomized
  trial)
• As the primary endpoint in phase II screening
  trials where a decision is being taken about
  future of drug or regimen

4
Why Measure Response?

• The word response is used in a number of
  contexts
• To describe outcomes in daily practice (my
  patient is responding to treatment)
• As a surrogate for benefit (e.g. in randomized
  trial)

• As the primary endpoint in phase II screening
  trials where a decision is being taken about
  future of drug or regimen. RECIST criteria
  developed for this

5
Response Criteria in Clinical Trials

• Clinical cancer research takes place in an
  international arena, thus we need a common,
  standard language to describe key methods and
  definitions for trial outcomes, such as
• Toxic effects terms and grades
• Time to event definitions
• Tumour response definitions

6
Response Criteria in Clinical Trials

• In early drug development
• Tumour shrinkage has long been used to provide a
  signal that new agents may be effective
• Anatomic-based criteria therefore required to
  describe and categorize patient outcomes
• WHO/others defined CR, PR, SD, PD
• To be useful as endpoint, Response
• Must be defined consistently
• Must be implemented consistently

7
Response Criteria in Clinical Trials

• Mid-late 1990s International working group began
  to meet to address some shortcomings of WHO. For
  example
• Complexity (bidimensional measurements)
• New technologies (CT)
• Silent on many areas so open to varying
  interpretation
• i.e. the standard was no longer standard

8
Response Evaluation Criteria in Solid Tumors
RECIST Working Group
EORTC P. Therasse M. Van Glabbeke (S) A.T.
van Oosterom J. Verweij NCI US S. Arbuck L.
Rubinstein (S) M. Christian R.
Kaplan NCICCTG E. Eisenhauer NDDO J.
Wanders UK S. Gwyther (R)

• 1995 International representation from different
  research organizations
• Revisit definitions, assumptions, implications
• Harmonize to the best standards
• Simplify where possible
• Update with new concepts
• An ongoing process

9
Response Evaluation Criteria in Solid Tumors
RECIST Working Group

• 1995 International representation from different
  research organizations
• Revisit definitions, assumptions, implications
• Harmonize to the best standards
• Simplify where possible
• Update with new concepts
• An ongoing process

10
RECIST Working Group 1995 - 99

• Consensus approach
• Reviewed different guidelines/criteria in use
• Changes if possible supported by data/literature
• First draft new criteria in 1997
• Consultation ICH approach
• US - Canada - Europe - Japan
• Industry - Regulatory - Research Groups
• International Workshop to discuss/resolve issues
  October 1998
• Presentation ASCO 1999, Publication 2000

11
(No Transcript)
12
Response Evaluation Criteria in Solid Tumors
(RECIST)P. Therasse et al, JNCI 2000

• Major application intended for trials where
  response is primary endpoint

13
Key RECIST Elements

• Unidimensional measurement of longest diameters
• Measurable lesion gt 20 mm (10 mm Spiral CT)
• Identify up to 10 measurable lesions maximum 5
  per organ. Follow sum of longest diameters (SLD)
• Response Categories
• PR 30 decrease in SLD compared to baseline
• PD 20 increase in SLD compared to lowest value
  on study
• CT scan preferred imaging modality. No ultrasound.

14
Since 2000

• Many publications comparing RECIST to WHO and
  others.
• With rare exceptions (e.g. mesothelioma)
  confirmed usefulness and reliability of
  unidimensional measurement
• A number of questions and issues have arisen

See review in Therasse et al. Eur J Cancer 2006
15
Issues Arising since RECIST Implementation

• Minimum number of lesions Can fewer than 10
  lesions be assessed?
• Use of RECIST in randomized trials
• Use of newer imaging technologies such as PET and
  MRI.
• Use of RECIST in trials of non-cytotoxic drugs.

16
Minimum Number of lesions

• 10 lesions (maximum 5 per organ site)
• Arbitrary choice. Unlikely to underestimate
  overall tumour burden.
• But not evidence based.
• What kind of evidence could lead to change?

17
Example 10 Lesions
18
Example 6 largest selected
19
Example 3 largest selected
20
How Many Lesions?

• Actual patient data analyzed as on previous
  slides will help determine what minimum number
  of lesions can be
• ? without changing the overall outcome or
  interpretation of a trial
• Ongoing project at EORTC data centre to examine
  this numerous trials and tumour measurement data
  from thousands of patients

21
RECIST in Randomized Clinical Trials (RCTs)

• Issues
• Progression-Free Survival or Time to Progression
  are increasingly common primary endpoints in RCTs
• Assessment of progression requires monitoring
  tumour size/number of lesions
• Because of the need to assess disease
  progression, RECIST use in RCTs has become common

22
RECIST in Randomized Trials

• This presents two issues
• Can rigorous response assessment requirements be
  relaxed?
• How to assess progression in patients who do not
  have measurable lesions?

23
Measuring Objective Response in Phase III Trials

• RECIST paper indicates that when response is not
  primary endpoint, modifications may be allowed.
  e.g.
• Fewer than 10 lesions
• No 4-week confirmation
• Such changes must be in protocol, not applied
  post-hoc.

24
RECIST in Randomized Trials

• Issues
• Can rigorous response assessment requirements be
  relaxed?
• How to assess progression in patients who do not
  have measurable lesions?

25
Progression in Phase III Trials

• Important issue, since as noted PFS and TTP
  becoming common primary endpoints.
• No problem if entry is restricted to patients
  with measurable lesions!
• But what about patients with non- measurable
  disease only?

26
Progression in Non-Measurable Disease

• Fundamental problem how to measure an increase
  in that which is not measurable?
• Options
• Count new disease only
• Look for unequivocal progression of
  non-measurable lesions subject to external
  review
• Something else?
• Go back to using overall survival?

27
Progression in Non-Measurable Disease (2)

• This is not a problem unique to RECIST!
• Need wider discussion on how to handle this with
  clinical researchers, regulatory officials and
  pharma.
• At minimum today, for critical (new drug
  approval) trials where PFS is primary endpoint,
  regulatory agencies will require external review
  of imaging before considering the trial results
  acceptable.

28
What about newer imaging technologies (PET/CT and
MRI)?

• Appendix I in RECIST paper
• In principle PET/CT and MRI may be used to assess
  SIZE provided they
• Can detect change in size from minimum baseline
  size that is representative of PR or PD
• e.g. if 10 mm lesion minimum at baseline, will
  modality detect 2-3 mm change?

29
What about Functional Changes?

• RECIST
• Based on anatomical tumour size
• Does not take into account phenomena such as
  tumour cavitation
• Does not take into account metabolic function
• Does not take into account blood flow parameters

30
Publications Proposed standards for
function/flow

• Young H, Baum R, Cremerius U, et al Measurement
  of clinical and subclinical tumour response using
  18F-fluorodeoxyglucose and positron emission
  tomography review and 1999 EORTC
  recommendations. European Organization for
  Research and Treatment of Cancer (EORTC) PET
  Study Group. Br J Cancer 1999 351773-82.
• Leach MO, Brindle KM, Evelhoch JL, et al
  Assessment of antiangiogenic and antivascular
  therapeutics using MRI recommendations for
  appropriate methodology for clinical trials.Br J
  Radiol 200376 Spec No 1S87-91.

31
Critical Question

• In terms of clinical benefit to an individual
  patient (survival, symptom improvement), or in
  terms of signaling that a new drug will improve
  survival .
• what is meaning of a change induced by a new
  drug in measures of tumour blood flow or
  metabolism?

32
Are changes in flow and function meaningful in
screening new drugs?

• To answer this we need to amass clinical data
  sets where
• Agents selected for development because of
  changes in these measures (no obj. response)
• And the results of randomized studies of these
  same drugs are known
• PTK787 story suggests more data needed before
  using these measures alone to screen new agents.

33
Are RECIST Applicable in Trials of Non-Cytotoxics?
34
Other proposed endpoints

• Stable disease above a minimum proportion and/or
  beyond a minimum duration
• Progression free survival
• Non-progression rates
• Tumour shrinkage lt PR

35
The Problem

• With the exception of the last proposal (ltPR
  shrinkage) all of the other endpoints are in fact
  RECIST categories!
• RECIST
• DOES provide categories for describing how
  patients tumours change in size
• DOES NOT state a minimum duration of stable
  disease (it DOES state it should be specified in
  each protocol)
• DOES NOT provide guidance about what proportion
  of patients within those categories in a trial
  signals a new drug may be active

36
So dont be FAZED

• One does not need new response criteria for
  assessing non-cytotoxics
• One may need to change the usual hypotheses
  that drive phase II design.so instead of a
  response rate of interest of 20 (typical for
  many cytotoxic phase II trials), we could
• Look for PR CR rate of 10
• Look for SD rate gt50
• Look for CR PR SD rate gt60
• (or whatever seems meaningful)

37
Phase II Results Targeted Drugs
38
RECIST in Trials of Targeted Drugs

• Major issue is regarding design (hypotheses,
  sample size, randomization etc.), not how one
  measures tumour size.

39
RECIST Revisited SUMMARY

• In clinical cancer research, particularly phase
  II screening trials, standard criteria to
  describe change in burden of tumour are needed.
• RECIST criteria widely adopted for this need.
• However, some issues identified that need further
  work/resolution
• Workload can same information be obtained by
  following fewer lesions?
• Assessing the role of functional imaging in
  screening new drugs
• Determining how to assess progression in phase
  III trials when non-measurable disease only is
  present

40
What is Happening Now?

• RECIST Working group has re-convened to examine
  these issues in systematic fashion
• Revised version of RECIST for assessment of
  anatomical tumour changes planned for 2008 will
  address several issues.
• Role of functional imaging in drug development
  requires more systematic evaluation, longer
  follow-up to clarify if it can complement of
  replace RECIST in future.

41
Thank you for your attention!