February 20, 2008
On Avastin, Why Not Register Results?
The Food and Drug Administration is considering expanding Avastin's labeling to include breast cancer. The outcome could have major policy implications for future cancer drug approval decisions, since the data submitted by Genentech in support of this anti-angiogenesis drug only shows an increase in time to tumor progression, not an overall survival benefit. The Wall Street Journal's Marilyn Chase this morning provides a good overview of the issues, including their financial implications for insurers, patients and Genentech.
A couple of things stand out. First, the drug's proponents claim that keeping the tumor in check by itself improves quality of life. Yet the side effects of adding Avastin to the traditional chemotherapy regime seem quite nasty. What's the evidence that quality of life was improved by increasing patients' time to tumor progression with this drug?
Moreover, if the data in these short-term trials don't show increased survival, then it is incumbent on the FDA to find out no matter how it decides the immediate issue of expanding the label. Thousands of women are already receiving this drug off-label. That would probably become near universal if it approves the drug for metastatic breast cancer, which kills 45,000 annually.
Traditionally, the FDA has asked companies to conduct post-approval clinical trials to resolve outstanding questions about the benefits and/or risks of a new drug. More often than not, the companies never complete them. In this case, Genentech apparently has one already underway, with data due in a few years.
But the agency could add on another approach. Why not require all physicians who use the drug, whether on or off-label, to keep careful (electronic) records of their patients' performance on the drug, and send that data to the FDA for analysis by the agency's drug evaluation and drug safety units? That data should also be open to outside investigators. (Think of what this would do to promote electronic medical record keeping, but that's another story.)
Observational data is not the same as data from a well-controlled, randomized clinical trial. But it seems to me that researchers would have a valuable new tool for evaluating drugs if the health care system began collecting solid epidemiological data from the real world use of drugs. Such studies would offer patients and their physicians another important source of information when deciding what is the best individual choice when it comes to imperfect drugs like Genentech's Avastin.
Posted by gooznews at February 20, 2008 08:33 AMWell I see the FDA OKed Avastin for breast cancer. As a supplier of research lab equipment- Genentech is our biggest customer - hooray. But as the husband of a breast cancer 'survivor' I scratch my head and wonder why the FDA did this? I suppose because progress fighting metastatic cancers has been so fruitless, the bar has been lowered again with 'success' now measured in a few months. Whatever happened to the concept of curing a disease? What a cruel hoax this is for these poor unlucky souls- but what a bonanza for Genentech.
Posted by: Doc Wrigley at February 22, 2008 04:55 PM Avastin neither lengthens life nor improves quality of life in metastatic cancer yet is "approved" by the FDA. Approved for what??? Apparently only to enrich the pocketbooks of Genentech who charges $100,000 year for a worthless drug. Unbelievable!!
The FDA "bar" is non-existent........as long as a drug can make money for Big Pharma and the Forced Drug Administration.........it will be approved. When will this evil corrupt behavior end? When will Americans wake up and put these white collar criminals in jail?
WHAT YOU ARE ADVOCATING IS WHAT PEDIATRIC ONCOLOGY HAS BEEN DOING FOR YEARS...AT LEAST FOR THE MAJORITY OF CHILDREN BEING TREATED OFF LABEL (WHICH IS BY DEFINITION SINCE ONCOLOGY DRUGS ARE NOT TESTED IN CHILDREN PRIOR TO APPROVAL). THAT COMMUNITY HAS MADE THE COMMITMENT TO LEARN WHAT WORKS AND WHAT DOESN'T -- WHILE AT THE SAME TIME TRYING TO MOVE THE TREATMENT ENVELOPE FOR INDIVIDUAL KIDS AS WELL AS THE ENTIRE POPULATION OF KIDS.
SOME OF US HAVE ADVOCATED FOR A "TREATMENT IND" TYPE ARRANGEMENT FOR SITUATIONS LIKE THIS - THAT IDEA IS VERY UNPOPULAR WITH SPONSORS (AND THE FDA SEEM UNINCLINED TO CONFRONT INDUSTRY ON THIS). THE OTHER POSSIBILITY, NOT THAT DISIMILAR TO A TREATMENT IND, IS TO HAVE A INTERMEDIATE CONDITIONAL APPROVAL STATUS WHERE A LIMITED POPULATION OF PATIENTS WOULD RECEIVE THE DRUG (BUT MORE THAN THE SMALL NUMBER IN PRE-APPROVAL TRIALS AND OVER LONGER PERIODS) AND THE KIND OF LEARNING YOU DESCRIBE WOULD TAKE PLACE. THIS WOULD BE A PRE-CONDITION OF MOVING TO FULL APPROVAL STATUS SHOULD THE RESULTS WARRANT IT. THUS THE CARROT AND THE STICK.
Posted by: ARTHUR LEVIN at February 24, 2008 11:39 AMThe idea of using real clinical experience of real doctors taking care of real patients using a new drug like avastin to learn how to better treat patients for a difficult to treat disease like metastatic breast cancer seems quaint and well intentioned but unscientific these days.
The notion that "observational data is not the same as data from a well controlled clinical trial." has been effectively used by drug companies like Genetech and regulatory agencies like the FDA remove the clinical experience and potential expertise of practicing doctors from the
equation of both drug approval and standards of care.
I would argue that the the data from well controlled clinical trials reflects the intent of the trial design and yields results that reflect that intent. Well controlled clinical trials do not intrinsically advance knowledge such as how to best treat difficult to treat patients where many potential factors can impact on clinical outcomes.
The idea that randomized clinical trials are intrinsically better than an observational data base as a way of advancing clinical knowledge needs to be critically examined. Whats needed is a strategic approach in which clinical experience and well designed clinical trials feed each other.
I totally agree with the idea to create a treatment registry to track clinical outcomes for patients with difficult to treat diseases. This would bring the doctor and patient back into the knowledge acquisition process... something very threatening to the narrow interests of drug companies like Genentech and their sponsored physician experts/Key Opinion Leaders who shape clinical practice and clinical research.
Its ironic that the best support for approving avastin rests on the hope that some doctors will figure out a way to use it in clinical practice that gives better results than the randomized clinical trial.
Something is very "rotten in the state of Denmark"
when billions of dollars can be extracted from our beleagured health care system now that the FDA has approved avastin on ther basis of the results of the randomized clinical trial that showed it delayed progression of breast cancer but didn't extend life.
There needs to be a mechanism requiring drug companies to support and pay for the kind pof treatment registries that would help doctors learn how to best use their drugs to treat their patients and that the often pharma supported patient centerd consumer groups need to start to demand this.