MULTIMODALITY CLINICAL TRIALS

1
MULTI-MODALITY CLINICAL TRIALS

• JOEL E. TEPPER, MD
• Department of Radiation Oncology
• University of North Carolina School of Medicine
• Chapel Hill, NC

2
Multi-Modality Clinical Trials

• Clinical and translational research are team
  sports
• Ability to do high quality research depends on
  the proper integration of all members of the
  research team
• Proper integration of people as well as concepts

3
UNC- NCAA champions-2005
4
Multi-Modality Clinical Trials

• Use of more than one modality as part of the
  clinical study
• Combination of surgery, radiation therapy and/or
  chemotherapy
• Combination of multiple chemotherapeutic agents
  or biologics
• Modifiers of toxicity

5
Multi-Modality Clinical Trials

• Modalities almost always interact at some level
• Biologic- enhancing cell killing
• Biologic- toxicity modification
• Patient selection- defining appropriate patient
  population
• Surgery
• Pathology
• Radiology
• Competing risks influencing the end point to be
  studied

6
Potential Problems

• Baseline response rate
• Baseline cure rate
• Baseline toxicity
• Interdependence of one modality on the other for
  therapeutic effect
• Competing risks for end-point of interest

7
Multi-Modality Clinical Trials

• Use GI cancer as the clinical model to illustrate
  some of the issues
• Radiation sensitization
• Surgical adjuvant therapy

8
Multi-Modality Clinical Trials

• Must consider whether one is trying to exploit an
  interaction between agents or evaluating two
  agents being used relatively independently
• Does one expect an interaction in either
  therapeutic or toxicity end-points
• What are the biologic parameters relating to that
  interaction

9
Phase I Study

• Conventionally evaluating tolerance to therapy
• Pre-existing toxicity from base-line therapy
  must be well defined
• Evaluate for increment in toxicity or tolerance
  of the overall approach
• Often hard to find a truly comparable base line
  study

10
Phase I Study

• Is one agent/approach more important than
  another?
• Should one start at a baseline dose of one agent
  and then escalate a second or use low initial
  doses of both agents and then escalate both?
• Escalation of 2 agents needs to be sequential
• Drug-radiation interaction
• Drug-drug interaction

11
Gemcitabine Radiation Sensitization

• Tolerance of upper abdomen to radiation therapy
  well defined
• Tolerance of systemic gemcitabine well defined
• Laboratory studies demonstrated major degree of
  radiation sensitization
• Need to define biological aim- emphasize
  systemic or radiation sensitization effect?
• Would result in different study designs
• Standard RT with escalating gemcitabine
• Standard gemcitabine with escalating dose of RT

12
Gemcitabine Radiation Sensitization

• UNC trial
• Locally advanced pancreatic cancer
• Standard radiation dose
• Twice weekly gemcitabine to try to maximize
  radiation sensitization- laboratory based
• Used very low initial gemcitabine dose because of
  laboratory data

13
Gemcitabine Radiation Sensitization

• Interaction between agents caused major toxicity
  in initial clinical studies
• Gemcitabine dosage with full dose RT is now 1/3
  or less of that used with weekly drug alone
• Sufficient to obtain sensitization
• More frequent scheduling of drug may be a better
  strategy with this combined modality approach

14
Phase II Study

• Conventionally evaluating for response of a new
  agent or new approach
• Combining two agents- usually a moderate response
  for the baseline agent
• Need to determine an estimate of the end-point of
  interest to the baseline therapy alone
• Design the study to develop a point estimate for
  the end-point of interest with combined modality
  therapy

15
Phase II Study

• When adding biologics to cytotoxics there may not
  be an expectation of alterin response rate
• 1 year survival estimate may be a better metric
• Studies will generally need to be larger to get a
  reasonable estimate of these end-points compared
  to single agent response rate

16
Phase II Study

• Adding more aggressive local treatment to
  patients with high risk of metastatic disease
  will likely not change early end-points
• Need to evaluate late end-points to have a
  reasonable chance of detecting a positive outcome

17
Phase II Study- End-Points

• Because late events may be of primary interest
  (and these may take too long), surrogate
  end-points often needed
• Biological
• Immunological
• Micro-array panels
• Functional imaging
• These approaches have often not been validated

18
Phase II Combined Modality- End-Points

• Clinical response rate
• Pathological complete response rate
• 1 or 2 year survival
• Time to progression

19
Phase II Design

• Adding a biologic modifier to a standard therapy
• If baseline response is very low, major biologic
  modification might not be detectable clinically
• If baseline response is very high, difficult
  statistically to detect the difference
• Moderate baseline response usually optimal

20
Phase II Study

• Toxicity could be a late event
• Major issue with radiation therapy dose
  escalation studies combined with chemotherapy
• Need to make reasonable compromises on avoiding
  major adverse events and allowing studies to move
  to completion
• Patients must be informed fully regarding risks

21
Phase II Strategy-Colon and Rectal Cancer

• Take high-risk patient subset
• Colon cancer s/p hepatic resection for liver
  metastases
• Treat post-resection with new agent and make
  point estimate of 1 or 2 year survival
• Local-regional esophageal cancer
• Add new agent to baseline chemoradiation protocol
  and estimate path CR rate or 1 and 2 year local
  control, and survival

22
Phase II Strategy

• Rectal cancer- radiation sensitization with
  oxaliplatin
• Laboratory data have demonstrated substantial
  radiation sensitization
• Timing of drug and radiation is likely important
• RT should be given after oxaliplatin
• Minimal clinical use of the combination

23
HT-29 Xenograft- Radiation /- oxaliplatin
24
RT oxaliplatin- rectal cancer

• Considerations in study design
• Assure that potentially curative surgery could be
  performed
• Maintain the known radiation/5-FU effect
• Maximize radiation sensitization with altered
  timing of oxaliplatin
• Define useful short term end-point

25
RT Oxaliplatin- rectal cancer

• Initiated study with T4 tumors
• Used standard RT/continuous infusion 5-FU
• Dose-escalated oxaliplatin using weekly
  oxaliplatin and starting at a dose lt1/2 the
  standard
• Ran as initial Phase I with immediate switch to
  Phase II after proper dose defined

26
RT oxaliplatin- rectal cancer

• Toxicity endpoints
• Acute toxicity
• Surgery done safely and on schedule
• High priority was not to compromise surgery in
  patients with curable disease
• Therapeutic endpoints
• Pathological complete response rate
• Based on estimate of 15 path CR with
  conventional therapy

27
RT oxaliplatin- rectal cancer

• Statistical considerations

28
RT Oxaliplatin- rectal cancer

• Assume path CR rate of 15 with conventional
  therapy
• Assume path CR rate of 30 with new regimen
• Accept proceeding if 5/25 (20) path CR is
  observed
• If true path CR rate is 30
• 5/25 observed CR will be seen 91 of time
• If true path CR rate is 10
• 5/25 observed CR will be observed 10 of the time

29
RT oxaliplatin- rectal cancer

• Initial dose- oxaliplatin- 20 mg/m2/wk
• Maximum dose- 50mg/m2/wk
• No undue acute or surgical toxicity
• Data on pathological complete response
  encouraging (Ryan et al, ASCO Proceedings, 2004)
• Regimen now being test in NSABP R04

30
(No Transcript)
31
(No Transcript)
32
Phase III Study

• Generally a major interdependence of modalities
• Quality control of each modality can be of
  enormous importance in defining whether a therapy
  should be used
• Produces an interaction which can effect the
  study outcome

33
Rectal Cancer Adjuvant Therapy

• Adjuvant chemoradiation therapy is used for T3
  and/or N disease
• Are all of these patients truly at high risk for
  failure?
• Could be dependent on quality of surgery and
  pathological evaluation

34
Rectal Cancer Adjuvant Therapy

• Importance of number of nodes
• Risk of failure dependent on node positivity
• Poor surgery could result in high local failure
  rate by inadequately treating the primary
• Poor surgery or poor pathology could result in
  spuriously high failure rates (local and distant)
  for low stage patients due to understaging

35
P0.003
36
Circumferential Margin and Local Recurrence

• 141 potentially curative resections
• Total local recurrence rate- 25
• 35/141 (25) had tumor within 1 mm of the
  circumferential margin
• Local recurrence- 78 with positive margin vs 10
  without
• Survival-24 with positive margin vs 74 without

Adam and Quirke- Lancet- 1994
37
MRI-Phased Array Coil
T3N1
Beets-Tan, Lancet-2001
38
Impact of surgeon

• Evidence that more experienced surgeons produce
  better local control
• 680 pts in 3 randomized trials
• Surgeons who operated on gt10 pts on study
• Improved local control- 17 vs 10 (plt0.005)

Stocchi, ASCO Proceedings- 1999
39
ADJUVANT THERAPY

• Including patients who are truly node positive in
  a cohort of patients staged as node negative can
  spuriously suggest an advantage to adjuvant
  therapy in node negative patients
• Do the results with adjuvant therapy apply to all
  surgeons?
• How would one make such a determination?
• Does one design a study for the average surgeon
  and pathologist or for the best?
• Results could be dramatically different

40
Postoperative Adjuvant Therapy

• Postoperative adjuvant radiation therapy is
  probably not required when certain criteria are
  met
• TME by an experienced surgeon
• Node negative after adequate nodal assessment
  (gt14 nodes in the specimen)
• Negative margins after formal evaluation- radial
  and distal
• High rectal tumor

41
COMBINED MODALITY THERAPY

• One must consider multiple issues in study design
• Biologic interaction between modalities
• Patient selection
• Quality control
• Base line data- response, toxicity, survival
• The same issues as in other studies,but
  approached differently

42
(No Transcript)