Conference on Clinical Cancer Research: Meeting Report

Nancy Roach is the Founder of C3 and Chair of the Board of Directors.

On September 14, I participated in the second Conference on Clinical Cancer Research, hosted by Friends of Cancer Research and the Engelberg Center for Health Reform at the Brookings Institute. The 2008 conference identified specific areas of clinical research that could be “de-bureaucracied”. In 2009, the conference presented the progress made in the past year, and identified next steps.

Following is background and a report on my panel’s discussion regarding centralized review of scans to determine progression free survival in phase 3 trials.

Warning: this meeting delved into highly technical aspects of clinical research, so the following is a little geeky!

Background

The Food and Drug Administration (FDA) regulates drugs in the United States. Before a cancer drug such as Avastin (bevacizumab) or Erbitux (cetuximab) can be sold to patients, the company making the drug must provide clinical trial results that show:

How the drug should be used – in what patients, at what dose?
Side effects, especially common or serious side effects
Impact – does the drug help patients live longer? Shrink tumors?

When a drug shows clear patient benefit and acceptable side effects, the FDA review process can be very straightforward. Gleevec (imatinib) for chronic myeloid leukemia (CML) patients is an example of this type of review. Patients in Gleevec clinical trials were not only living longer, but also many of them showed signs that their CML had gone away. FDA approved Gleevec in 2 months, one of the shortest reviews on record.

Unfortunately, many cancer drugs don’t show such striking results. The clinical trial data taken to FDA might show that the drug makes tumors shrink (response rate) or keeps tumors from growing for a longer time (progression free survival), but not that the drug helps patients live longer (overall survival). Ultimately, increasing survival or improving quality of life while patients are alive is the goal. When FDA reviews data, they frequently grapple with questions like:

If a drug increases progression free survival but not overall survival, and if the drug has significant side effects, how do patients benefit?
If overall survival is increased by a few weeks, do patients benefit?
If trial results show tumor shrinkage, were the scans that show the shrinkage read correctly?

FDA works through these questions in private meetings with sponsors (generally drug companies). FDA also has the ability to convene an official Oncology Drug Advisory Committee meeting where they can raise these issues publicly, get advice from ODAC members and comment from the public.

Goal of the Conference

The conference convenes groups that “dive into the weeds” of drug review, and look for ways to improve the process. This year’s conference focused on four specific areas:

How much side effect (safety) data needs to be collected in clinical trials for drugs that are already used to treat different forms of cancer?
If a trial is using scans to document tumor growth (progression), do all of the scans need to be double-checked by a blinded central review group?
As research finds biomarkers that predict whether a patient will respond to a drug, how can the tests identifying the biomarkers be effectively regulated?
As combinations of targeted treatments are tested in clinical research, what guidelines do companies need to ensure that the right data are produced for FDA review?

Panel 2: Local Evaluation vs Blinded Central Review of Scans, and PFS

I represented patients on the 2^nd panel, which is looking at ways to manage the review of scans. Companies bringing new drugs to market have blinded centralized review (BCR) of all scans in phase 3 trials. BCR is very expensive, and the question we considered is: Does BCR result in better data? Is it possible to audit a specific proportion of scans and perform BCR only if warranted? The panel was chaired by Dan Sargent from the Mayo Clinic, a world-renowned clinical trial statistician and member of C3’s Medical Review Network. The other panelists were:

Will Bushnell, GlaxoSmithKline
Ohad Amit, GlaxoSmithKline
Lori Dodd, National Institute of Allergy and Infectious Diseases
Richard Pazdur, FDA

Will, Ohad and Lori presented a couple of different ways to approach BCR auditing, rather than 100% BCR. The slide presentations can be viewed here. At that point, I was asked to present the “patient perspective”, which I did without slides.

Representing the patient perspective on these “in the weeds” issues can be challenging. Patients want drugs that work. Most aren’t wondering if a drug was approved because it increased overall survival or progression free survival. I raised these questions with C3’s Research Advocacy Listserv, a group of advocates involved with research projects, and used their feedback to generate my remarks, which follow in outline form:

We support pilot projects to test the auditing proposals. At the same time, we need to maintain a healthy skepticism about auditing vs 100% BCR until we see data.
Putting on a consumer hat, we have some questions:
- Progression free survival (PFS) can be a meaningful benefit to patients when the increase is ‘significant’ and data exists that shows PFS results in longer survival or improvement in quality of life. When PFS provides questionable benefit rather than meaningful benefit – for example, if the increase in PFS is a few weeks, and quality of life does not improve – should auditing be handled differently?
- We all agree that auditing is appropriate only in blinded trials. While some trials are technically blinded, they may not be blinded to patients and doctors because of different side effects. For example, if the new drug causes a rash, the local site will know which patients are on the trial drug. How does that effect auditing?
- The auditing proposals relay on a “large” treatment effect. In that situation, who defines “large”? Will it be defined prospectively or retrospectively?
Our perspective is cautious and conservative. This is because when Sponsors and FDA are negotiating these point behind closed doors, the bar for approval must be set high. Patients need drugs that provide significant benefit.
Conclusion: Let’s test the auditing proposals. If they hold, let’s put together solid regulatory standards that maintain a high bar for approval.

Richard Pazdur from FDA closed out our panel. He emphasized FDA concerns about the use of PFS as a surrogate endpoint, and suggested that FDA would hold an advisory committee meeting to clarify FDA’s position on appropriate use of PFS and the role of auditing.

C3 appreciated the opportunity to be involved with this discussion. We recognize that the drug development process is not “one-size-fits-all”, and work in many forums to streamline the process for the development of great drugs.

I’ll end with two quotes from our research advocates which I thought captured the essence of the patient perspective:

There are reasons that local investigators might want to judge progression as happening or not happening and be biased. They might want to enable a patient to continue on an experimental drug, or they might want to discontinue that individual’s treatment so they could cross over to what seems to the investigator to be the more promising arm.

To me, every scan should be verified because from the patient’s point of view, that individual may be continued in a trial or dropped from active treatment based on the scan. Whether research results can be considered valid if only a percentage of scans are verified is a statistical question. However, behind that black, white, and very gray image is an individual person whose future depends on an unbiased reading.

This news article was originally posted on October 7th, 2009 and was accurate at the time of publication. Since then, information may have changed or links may now be outdated. Please call our Answer Line 1-877-427-2111 for the latest information, or talk to your doctor before making any medical decisions.

Posted by Nancy Roach on October 7th, 2009