Title: Synopsis of FDA Colorectal Cancer Endpoints Workshop
1Synopsis of FDA Colorectal Cancer Endpoints
Workshop
- Michael J. OConnell, MD
- Director, Allegheny Cancer Center
- Associate Chairman, NSABP
- Pittsburgh, PA
2FDA Colorectal Cancer Workshop (11/12/03)
- Purpose
- Discussion of positive and negative aspects of
various endpoints for approval of new drugs for
colorectal cancer - Identify areas for further research
- Provide information to ODAC for recommendations
to FDA
3Colorectal Endpoints Workshop Format
- Regulatory background and summary of previous
approvals provided by FDA - Five presentations
- Interactive discussion by speakers and
multidisciplinary panel - Discussion of questions posed by FDA
4Goals of This Presentation
- Capsule summary of presentations and main points
of discussion - Focus on possible new endpoints for regulatory
approval of drugs for colorectal cancer - Help ODAC advise FDA
5Biomarkers or QOL in CRC Drug Approvals(Dr.
Charles Blanke)
- Not possible to consistently predict clinical
benefit based on reduction of CEA - ASCO guidelines do not recommend other biomarkers
for CRC
6Biomarkers or QOL in CRC Drug Approvals(Dr.
Charles Blanke)
- Methodologic issues of paramount importance if
QOL used as endpoint - Unknown whether changes in QOL reliably occur
with effective chemo - Cannot discriminate between safety and efficacy
- Best use of resources in colon cancer?
7Biomarkers or QOL in CRC Drug Approvals(Dr.
Charles Blanke)
- Clinical Benefit Response (Pain, performance
status, weight loss) - CBR does not adequately encompass symptoms
experienced by patients - Methodologic issues in assessment
- Not useful if asymptomatic (colon vs rectum)
8Endpoints in Neoadjuvant and Adjuvant Rectal
Cancer(Dr. Meg Mooney)
- Locoregional failures are usually symptomatic in
rectal cancer - Local tumor control at 3 years is an appropriate
endpoint for full approval - Pathologic complete response (pCR) -quality
control issues - Colostomy-free survival applies mainly to low
lying tumors
9Surrogate Endpoints and Non-Inferiority
Trials(Dr. Thomas Fleming)
- Primary endpoints sensitive, measurable, and
clinically relevant (eg. Survival decreased
symptoms) - Surrogate endpoints
- May reflect biological activity without
establishing clinical efficacy - Meta-analyses required to validate
- Validated surrogate endpoints are rare
- FDA has granted approval using surrogate
endpoints not formally validated
10Surrogate Endpoints and Non-Inferiority
Trials(Dr. Thomas Fleming)
- Non-inferiority trials
- Insufficient for curves to overlap
- Conservative margins to exclude significant
decrease in efficacy - Rigorous study conduct to avoid incorrect
conclusion of non-inferiority - Will results move the field forward (eg
significant decrease in toxicity)?
11TTP Clinical Benefit Endpoint in 1st Line
MCRC(Dr. Langdon Miller)
- Multiple effective therapies for metastatic
colorectal cancer have confounded the
relationship between early tumor control and
survival - Evaluation of symptoms problematic as endpoint
because progression frequently not symptomatic,
is subjective, and difficult to measure
12TTP Clinical Benefit Endpoint in 1st Line
MCRC(Dr. Langdon Miller)
- Arguments for TTP as endpoint for full approval
- Directly evaluates changes in disease burden
- Correlates with other outcomes (in particular,
survival) - Not confounded by subsequent therapies
- Offers utility as an endpoint in non-inferiority
trials (more rapid completion)
13TTP Clinical Benefit Endpoint in 1st Line
MCRC(Dr. Langdon Miller)
- Can be objectively quantified, reviewed, and
audited - Offers clear interpretation and straightforward
analysis - Conserves patient resources and hastens drug
development
14TTP Clinical Benefit Endpoint in 1st Line
MCRC(Dr. Langdon Miller)
- Correlation of TTP and Survival Was Highly
Significant two examples - 1000 patients in two phase III trials with
primary patient data - Meta-analysis on published summary results of 29
trials involving 13,000 patients
15TTP Clinical Benefit Endpoint in 1st Line
MCRC(Questions and Comments)
- Need for objective and reliable methodology for
assessing TTP - Does TTP reflect clinical benefit in its own
right? (Full approval) - Is TTP reasonably likely to predict clinical
benefit? (Accelerated approval) - RR, survival and toxicity also important
163-YR DFS as Endpoint in Adjuvant Colon Cancer
(Dr. Dan Sargent)
- Preliminary findings of meta-analysis to
determine if 3-YR DFS can replace 5-YR OS as
endpoint for colon cancer adjuvant trials - 12 clinical trials
- 38 treatment arms
- gt 10,000 patients
173-YR DFS as Endpoint in Adjuvant Colon Cancer
(Dr. Dan Sargent)
- Preliminary Conclusions
- 3-YR DFS seems to be an excellent predictor of
5-YR OS - Event rates were virtually identical (no impact
on sample size) - 3-YR DFS may slightly overestimate differences in
5-YR OS - Three studies significant difference in 3-YR
DFS (borderline p values) but no significant
difference in 5-YR OS - Not a formally validated surrogate
183-YR DFS as Endpoint in Adjuvant Colon Cancer
(Questions and Comments)
- Work in progress updated analysis today!
- Does improvement in 3-YR DFS represent clinical
benefit in its own right? - DFS is used for full approval in breast cancer
adjuvant therapy
19Key Questions for ODAC
- Should the following endpoints be recommended to
FDA for new drugs in colorectal cancer? - Full or Accelerated approval?
- Setting Endpoint
- Colon adjuvant 3-YR DFS
- 1st line metastatic TTP
- Rectal adjuvant 3-YR local control