December 11, 2008
Barbara Brenner, Former Executive Director
This morning’s session had two parts: a plenary session on combining hormonal therapies with “signal transduction inhibitors” (STI’s), and a general session on how to sequence the hormonal treatments of both tamoxifen and aromatase inhibitors (AI’s).
(See the slides of this presentation here and click on Thursday, December 11, Plenary Lecture 1.)
STI’s are drugs like Herceptin (trastuzumab) that inhibit signals between cells that are involved in the cancer promotion process. The theory of this morning’s presentation by Stephen Johnson is that some patients will develop a resistance to tamoxifen that can be overcome or prevented by giving an additional STI drug.
The need for strategies to overcome resistance has to do with the fact that some women treated with tamoxifen nonetheless develop recurrences or metastases, indicating that the drug has stopped working. As Dr. Johnson put it, “we give tamoxifen for 5 years and hope that the patient is cured.”
It’s important to note that there appears to be no cancer drugs to which some women don’t develop resistance. There have been sessions at this meeting on how to overcome Herceptin resistance.
Dr. Johnson pointed out that there are many unknowns about this approach: Who will benefit? Should the strategy be to prevent resistance or to overcome it once it arises? Are there biomarkers that might tell us which women will benefit?
In order to answer these questions, Dr. Johnson says it will be necessary to do biopsies to understand what cell pathways are active in a given patient so that the correct STI can be chosen to inhibit that pathway. He also says that the clinical trials to study this combination will have to take into account this need to be clear in the selection of patients.
I came away from this talk thinking about the difference between what is said in conferences like these, and what happens in clinical practice in communities outside the halls of science. While Dr. Johnson points out that more work needs to be done, what the doctors here heard was that if you combine an STI with a hormonal therapy, you can overcome resistance.
It’s still unclear who is likely to develop resistance. Will the doctors who read about this presentation decide to give both drugs to everyone so that they reduce the risk in the ones who will benefit? Will the result be yet more treatment—and more expense and more unnecessary side effects as we grasp for anything that will keep patients alive?
Not long after AI’s came on the scene for adjuvant treatment of breast cancer to reduce risk of recurrence, the medical community began to ask whether patients already on tamoxifen would do better by being switched onto an AI. And trials have been reported at this meeting before looking at the benefits of switching.
We heard 5 different presentations on the topic of which hormonal drugs, and in what order, and does it matter: Abstracts 11 (lasofoxifene [SERM] to reduce risk of breast cancer, 12 (meta-analysis of switching studies of tamoxifen and AI’s), 13 (letrozole and tamoxifen, alone and sequenced), 14 (tamoxifen alone v. tamoxifen switched to anastrozole), and 15 (tamoxifen v. tamoxifen switched to exemestane).
As just reading these brief descriptions makes clear, this is all enough to make anyone’s head spin. Each presentation had a take-home message, though, and those are easy to summarize:
Lasofoxifene (According to Wikepedia, lasofoxifene is a SERM—Selected Estrogen Receptor Modulator, like tamoxifen and raloxifene—under development by Pfizer for prevention of osteoporosis.) is a potent drug for women with osteoporosis that lowers the risk of breast cancer by 85%. Of course, this is a relative risk reduction number, which is meaningless to individual patients. The number of cases of breast cancer in this study of over 8500 women was quite small (15 in the lasofoxifene group, 21 in the placebo group), so this is another situation of giving lots of people a drug that comes with an increased risk of deep vein thrombosis to reduce the risk of breast cancer in a few.
The drug is not on the market—it hasn’t been considered yet by the FDA. And the long financial disclosure slide that led this presentation showed that most of the researchers are in some way or another on the Pfizer payroll.
The meta-analysis of the switching studies showed that AI’s reduce the rates of recurrence when compared to tamoxifen, and switching to an AI from tamoxifen provides a statistically significant improvement in overall survival. Keep in mind that statistically significant does not mean large or meaningful. What it means is that the result is real, not a chance outcome.
In the letrozole v. tamoxifen study, the monotherapy (no switching) comparison led to un-blinding of the trial after two years and cross-over so that about one third of the women who had been on tamoxifen started taking letrozole when the early results showed (1) improved disease free survival, (2) time to distant recurrence and (3) overall survival on letrozole compared to tamoxifen. The analysis of the switching arms of the trial, which compared starting on tamoxifen v. starting on letrozole, showed no statistically significant differences in outcome. Despite this fact, the presenter urged starting patients on letrozole because of a trend in the data in favor of letrozole first.
The biggest problem with this trial is the un-blinding and cross-over that occurred, and that is becoming increasingly common in clinical trials. As the researcher and others noted, cross-over seriously complicates the analysis of results, and very likely biases the results in favor of AI’s. While the reason for allowing cross-over is compelling—if something seems to work, shouldn’t everyone have the chance to get it?—the practice undermines the results of the trials. Wouldn’t women be willing to forego un-blinding and cross-over if they understood the need to have more reliable trial results? Shouldn’t we ask them?
The tamoxifen alone v. tamoxifen switched-to-anastrozole was another trial in which early results led to un-blinding and cross-over, which again inhibited the analysis of the results. Nonetheless, the researcher concluded that switching to anastrozole significantly (in a statistical sense) improved relapse free survival, as well as overall survival. Patients who switched had more fractures and more neurological problems than those who didn’t. Also, higher levels of ER (estrogen receptor) and PR (progesterone receptor) were predictive for longer relapse free and overall survival. This trial was paid for by Astra-Zeneca, the company that held the patent on tamoxifen until it expired, and currently holds the anastrozole patent.
In the tamoxifen v. tamoxifen switched to exemestane trial (called TEAM), disease free survival was the primary endpoint, evaluated at 2.75 years of follow up (before any patients were switched). Exemestane’s impact on disease free survival was statistically significant. Hypertension was worse on exemestane, as were osteoporosis and athralgias. Vaginal discharge and thromboembolism were greater on tamoxifen. This was an open-label trial—another practice that is increasing common in cancer clinical trials. It means that the doctors know what drug they were giving, and the patients knew what drug they are taking. It’s impossible to know how this knowledge affects either the doctors perceptions of how the patients are doing, or the patients’ reports of their side effects. But what is known is that in these trials, as in general clinical practice, many patients—29% of tamoxifen patients and 18.9% of exemestane patients—discontinued treatment.
So, what’s the take home for all of these studies? It seems to be that the drum beat for AI’s continues to build, though which one to give, whether to give tamoxifen first, and how long to continue treatment continue to be unanswered questions. We have more information, but not more knowledge.
Testing for CYP2D6
Sometimes I hear something at this meeting because I happen to be at the right place at the right time, and not by plan. I walked in a little early for a presentation I wanted to hear, and heard a scientist finishing a talk in which he advised that patients be tested for CYP2D6 to determine their ability to metabolize (and therefore benefit from) tamoxifen. Another situation where individualizing treatment may be pointing to what treatments people don’t need.
Genetic Susceptibility: Smaller and Smaller
In the presentation by the AACR (American Association for Cancer Research) Outstanding Investigator Award, Douglas Easton discussed the search for more common genes that confer some increased risk, but not a lot of increased risk. This refers to the search for genes to explain the breast cancers that are not explained by the BRCA1 and 2 genes, which make up only 5% to 10% of cases.
Researchers are now focused on SNP’s—single nucleotide polymorphisms, the DNA variations within a single gene neucleotide (CATG). There are many thousands of these variations, and technology is allowing researchers to analyze and start to understand them.
Easton acknowledged that we don’t yet know what the SNP differences mean. More importantly, he noted that what is being found does not bear out the early guesses by genetic researchers of what would be found. First it was genes that would answer all our questions. Then it was the proteins expressed the genes. Now it’s SNP’s. We look at smaller and smaller parts, and we don’t seem to understand much more. Maybe we should start looking at bigger things.
Risk Assessment and Risk Reduction: Nothing New
There was a mini-symposium on this topic, but there were no new ideas on how to manage patients at high risk of breast cancer. The most interesting statement made in a presentation on “diagnosis and prevention” in women at high risk was made about different kinds of mastectomy, and the fact that mastectomy results in irreversible sensory loss. As the presenter William Wood noted, patients don’t know this if they read the mainstream media on breast cancer. The example he gave was Cosmo.
So, no big news here as far I can tell, at least not yet