Skip to content | Skip to institutional links

Common menu bar links

Drugs and Health Products

Print |

Need Larger Text?

Record of Decisions: February 26, 2009, Scientific Advisory Committee on Oncology Therapies (SAC-OT)

February 26, 2009

Help on accessing alternative formats, such as Portable Document Format (PDF), Microsoft Word and PowerPoint (PPT) files, can be obtained in the alternate format help section.

(PDF Version - 36 K)

Contact: Bureau of Policy, Science and International Programs Enquiries

Record of Proceedings

Committee Members Present: Alexander Paterson (Chair), Lynne Nakashima, Wojciech Morzycki, Harley Ast, Kong Khoo, Ron MacCormick, Diana Ermel, Ian Tannock, Lesley Seymour, Allan Donner

Regrets: Rick Abbott, Joseph Ayoub, Charles Butts, Alwin Jeyakumar, Kara Laing, Mark Levine, Maureen Trudeau

Health Canada Representatives Present:
Barbara Benning, Thea Mueller, Bob Li, Larissa Lefebvre, Jackie Moore, Andrew Raven, Barbara Rotter, Martha Harczy, Karen Johns, Aleksandr Gamarian, André-Marie Leroux, Andrea Bell, Jean-Francois Ethier, Helen Mao, Karen Veltri, Kathy Soltys, Tatiana Lejen, Jennifer Kent, Elizabeth Prosen, Yuri Chenyakin, Tamara Ruby, Svetlana Spasova, Martin Szumski, Kate Van Der Giessen, Celia Lourenco, Roxanna Alexa, Nicholas Argent

Acronyms used in this record:
BMORS - Bureau of Metabolism, Oncology and Reproductive Sciences
COI - Conflict of Interest
HR - Hazard ratio
MRCC - Metastatic renal cell carcinoma
NOC - Notice of Compliance
NSCLC - Non small cell lung cancer
OS - Overall Survival
PFS - Progression-free survival
QOL - Quality of Life
RCT - Randomized controlled trial
SAC-OT - Scientific Advisory Committee on Oncology Therapies
TPD - Therapeutic Products Directorate

This meeting of the Scientific Advisory Committee on Oncology Therapies (SAC-OT) was called in order to discuss outstanding issues with the use of progression-free survival (PFS) as an endpoint in various forms of cancer. Questions and background documentation on this issue were formulated by Health Canada and sent to members prior to the meeting.

Opening Remarks (B. Benning, Therapeutic Products Directorate [TPD])

The TPD's Senior Executive Director opened the meeting, thanking the Committee members for their assistance in helping Health Canada to tackle the difficult regulatory issues associated with PFS.

Introduction and Conflict of Interest (COI) Declaration

The Chair led the Committee in round-table introductions and conflict of interest declarations, encouraging a free and open discussion. Upon completion of the declarations, it was unanimously agreed that all members could participate fully in the meeting.

Overview of Agenda

The Chair gave an overview of the day's agenda, introducing a new meeting format, which would allow the Committee to deliberate and formulate direct responses to each of the questions posed by Health Canada.

Health Canada Presentation: Part 1: Progression-Free Survival - A Follow-up (M. Harczy, Bureau of Metabolism, Oncology and Reproductive Sciences (BMORS))
*This presentation is available upon request.

Question and Answer Teleconference with Dr. Jennifer Knox

Dr. J. Knox, author of Progression-free survival as endpoint in metastatic RCC? (Lancet 2008;372:427-9) joined the Committee via teleconference for half an hour to act as an expert resource and answer questions from the Committee and Health Canada concerning her publication.

Health Canada Presentation: Part 2: Progression-Free Survival as an Endpoint in Metastatic Renal Cell Carcinoma (K. Johns, BMORS)
*This presentation is available upon request

Health Canada's Questions and SAC-OT Responses

Health Canada's questions and the Committee's responses are outlined below.

Part 1: Progression-Free Survival (PFS) versus Overall Survival (OS)

Question 1

Based on validation exercises and their outcomes, is PFS a surrogate of OS in the following cancers?
- Colorectal?
For metastatic disease, yes, but further evidence of clinical benefit is required. The magnitude of the PFS is critical. Where the magnitude of the improvement in PFS is small, further evidence of clinical benefit is essential. The effect of the treatment should be biologically plausible and consistent across trials. The PFS data should be of high quality (for example, externally validated and/or a properly controlled trial).
- Breast?
The evidence for validation of PFS as a surrogate marker for OS in breast cancer is less persuasive than for colorectal cancer. Where the magnitude of the improvement in PFS is small, further evidence of clinical benefit is essential. The effect of the treatment should be biologically plausible and consistent across trials. The PFS data should be of high quality (for example, externally validated and/or a properly controlled trial).
Should PFS be validated for common cancers only?

Current evidence has shown that PFS is, at best, an imperfect surrogate for OS. When used to consider drug approval, it needs to be considered in light of other attributes as noted above.
Can PFS be validated in rare cancers?

No, as randomized controlled trials (RCT) are unlikely to be feasible. With rare exceptions, PFS alone using non-randomized controls is not helpful in assessing approval for therapies in rare cancers. Other parameters such as response rate, symptom control and overall clinical benefit should be considered.
Are non-randomized Phase II PFS data meaningful?

No. See (c) above.
Should drug approval precede or follow validation?

Ideally, approval should follow validation, but in many instances this is unlikely to be available. Each application should be judged individually. In the case of a Notice of Compliance (NOC) based on PFS, the final safety, clinical benefit and mature survival data must be submitted in a reasonable timeframe after approval of the NOC.

Question 2

Does PFS constitute a direct measure of clinical benefit?

PFS is not a direct measure of clinical benefit but may correlate with symptom relief, which is a direct measure of clinical benefit.
If the magnitude of the PFS is large and there is minimal toxicity, there is likely to be a clinical benefit. If the PFS is small and there is significant toxicity, there may be little clinical benefit.
How would you define clinical benefit?

Evidence of a meaningful survival gain and/or symptom palliation and/or beneficial effect on quality of life (QOL). These benefits should outweigh the toxicity of the medication.
Does PFS address symptom control?

No.

Question 3

Is additional evidence for palliation required, aside from statistically significant PFS data?

If there is only a small PFS benefit, some additional evidence of clinical benefit is required.

Question 4

If Health Canada makes the decision to approve an oncology therapeutic product based on results for PFS, what type of information is useful to you as a prescriber?

Lack of data on OS?

Yes
Lack of data on symptom control?

Yes
No difference in Quality of Life (QOL)?

Yes

This information should be included in all three parts of the product monograph.

Question 5

For a targeted agent indicated for a common type of cancer, can a significant improvement in PFS be the basis for full approval?

For a targeted agent, an improvement in PFS alone may be the basis for full approval provided that the difference is statistically significant and clinically meaningful, considering unmet needs for the disease area. Supportive information on symptom and/or QOL improvement will help the application.

Many targeted agents act on a subset of patients. Predictive markers for efficacy and toxicity should be explored.

Part 2: PFS versus OS in Metastatic Renal Cell Cancer (MRCC) - A Case Study

Question 1

Is the analysis of the data presented in the Knox article sufficiently robust to conclude that PFS has been validated as a surrogate marker for OS and, thus, indicative of clinical benefit (based on surrogacy) in MRCC?

If yes, please indicate why.

If no, please indicate why and indicate what type of data and/or analyses are required in order to be able to conclude that PFS is a valid surrogate for OS in MRCC?

No. Due to confounding factors such as crossover, it is unlikely that PFS can be validated.
In MRCC, PFS can be regarded as a useful endpoint provided that the effect is large and/or if it is supported by evidence of improvement in symptoms and/or QOL. Consideration should be given to biological plausibility, consistency across trials and fulfillment of an unmet need.
Are patient-level data required or are study-level results sufficient?

If this were to be done, patient-level data are recommended.

Question 2

Can PFS be considered a direct measure of clinical benefit in MRCC? Given that it has been suggested that a substantial magnitude of PFS improvement is a clinical benefit to patients, in your opinion, do the data on PFS from the TARGET, RECORD and sunitinib versus interferon-alpha trials demonstrate a clinical benefit in MRCC? Yes/No - please explain why.

No, PFS is not of itself a direct measure of clinical benefit in MRCC, however, the magnitude of the benefit in the sunitinib trial, combined with supporting data of OS, response rate and QOL make it likely that it indicates true clinical benefit. In the other two trials, the hazard ratio (HR) for PFS is also large and they are addressing an unmet need, however, the supportive data, including data from other studies, are not as compelling.

Overall Conclusions

In summary, PFS will be at best an imperfect surrogate marker for OS, but if the magnitude of the benefit is large in the setting of the stage and type of malignancy under consideration, it may of itself indicate a clinical benefit.
*The Committee agrees that a change in terminology from PFS to progression free interval may be preferable in the future.